I

H
K2

3tolm.

^irams
TME CUSTODY OF THE BOSTON PUBLIC LIBRARY.
iELF N
!

FN909 5.12,37: 150

Digitized by the Internet Archive

in

2011

http://www.archive.org/details/treatiseofalgebrOOmacl

TREATISE
F

ALGEBRA,
1

THREE PARTS.
CONTAINING
I.

The fundamental Rules and Operations.

II.

The Composition and Resolution of Equations of all Degrees and the

;

different
III.

Affections of

their

Roots.

The Application of Algebra and Geo-

metry

to each other.

To

which

is

added an

APPENDIX,
Concerning the general Properties of

Geometrical
By
Late ProfefTor of

Lines.

COLIN MACLAURIN,
Mathematicks
the

M. A.

in the Univerfity of

Edinburgh, and Fellow of

Royal Society.

The

SECOND EDITION.

LONDON,
Printed
for

A.

Millar

and J.

Nourse,

oppollte

Katherine- Street in the Strand,

MDCCLVI.

ADAMS

T O

THOMAS,
LORD ARCHBISHOP
O
F

HIS

GRACE

CANTERBURY,
Primate of
As
a teftimony of

all

England

Gratitude

for the

friendfhip and generous protection, with

which His

Grace was

pleafed to honour

my

deceafed hufband,

THIS

TREATISE
IS

INSCRIBED,

BY HIS GRACE'S MOST OBLIGED

AND OBEDIENT HUMBLE SERVANT,

ANNE MACLAURIN;

TO THE

READER.
}@0g09()(

HAT' Mr. Macl aurin

_)Vtfri tfgtf,

had,

many

M5CM^C
letter,

/{/*

intended to publijh a Trea-

JubjeB, appears from a

to his

dated April ig, 1729,

honoured

friend

Martin Folkes,
the

JE/^j

now Prefiin

dent of the Royal Society *.

one of his Manufcripts,

And we find,

Plan of fuch a
with

Work, agreeing,
the contents

almojl in every article\

of

this

Volume

Had
his

the celebrated
his

Author lived to publijh

alone,

own Work,
:

name would,
it

have

been fufficient to recommend

the Publick

to the notice oj

but that I'ajk having, by his

lamented premature death, devolved to the gentlemen

the

whom

he left entrujled with his Papers,

Reader may reafonably expecJ fome account of the materials of which it confifts, a?id of
the care that has been taken in collecting
difpofing them, Jo as
thor's intention,
defigned.
* Phil. Tranf.

and

bejl to

anjwer the Authe

and fill up

Plan he had

N 408.
?

He

To
He feems,
had
thefe

the

Reader.

in compofing this Treatife, to have

three Objects in view.

i.

To give

the

general Principles and Rules

the Jl:ortejl, and, at the fame

of the Science, in
that

time, the moji clear

and comprehenfive manner

Agreeable
to this,

was

pofjible.
is

though

every Ride

properly exemplified, yet he does

not launch out into

what we may

call,

a Tauto-

He rejects fome applicalogy of examples. tions of Algebra, that are commonly to be met
with in other writers
-,

becaufe the
:

number of
by

fuch applications

is endlefs

and, however ufethey caiinot,

ful

they

may

be in Practice,

the rules

of good method, have place in an

elementary Treatife.

He

has likewife omitted

the Algebraical folution of particular

Geo-

metrical problems, as requiring the knowledge

of the Elements of Geometry thofe of Algebra ought to be

really are, entirely difiinc~i
felf to
;

from which
kept,

as they

referving to himrelation

treat of the

mutual

of the two

Sciences in his
rally fill,
too,

Third

Part,

and, ?nore gene-

in the

Appendix.

He
was

might think
the
lefs

that fuch an application

that Sir

ne-

ceffary,

Isaac Newton'*
is

excellent

Collect ion

of Examples

in every body's hands ,

a?id that there are

few Mathematical writers>

who

To
who
kind.

the

Reader.
nwnbers

do

not furnijh

of the fame

2.

Sir

Isa^c Newton'^

Rules,

in his

Arithmetica Univerfalis, concerning the Refolution

tions

of the higher equations, and the Affec-

delivered

of their roots, being, for the moji part, without any demonftrat ion, Mr*

Maclaurin
tife JJjou/d

had

defigned,

that his 'Trea-

ferve as a Commentary on that

For we here find all tkofe difncult pafjages in Sir Isaac'j Book, which have fo
Work,
long perplexed the Students of Algebra, clearly

explained and demonjlrated.

How
,

much fuch
learn

a Commentary was

w anted

we may

from

the

" The

words of a late eminent Author *. ableft Mathematicians of the laft are

"
<c

(fays he) did not difdain to write

Notes on

the

Geometry of Des Cartes

and

fiire-

"

ly Sir Isaac Nev/ton's Arithmecick no " lefs deferves that honour. To excite feme " one of the many fkilful Hands that our " times afford to undertake this Work, and
<c

to fhew the Neceflity of it, I give this " Specimen^ in an explication of two paf-

* sGrarjefar.de in
Arith. Univerf.

Prsefat.

ad Specimen Comment, in

(t

fages

To
u
{

the

Reader.
;

fages *

of the Arithmetica Univer/a/is

" which, however, are not the

cult in that

mod

difll-

Book."
learned
at
Profeffor fo
earnejlly
laji fee executed-,

What

this

wijhed for,

we

not fepa-

rately, nor in the loofe dijagree able form

which
every
the

fuch Commentaries generally take, but in a

manner

equally natural

and convenient

-,

Demonfiration being aptly inserted into

Body of
rable

the

Member

Work, as a uecefary and irfepaan Advantage which\ with ;

obvious enough to an attentive
it is

fome others,
Reader, will,

hoped, diftinguiflo this

Per-

formance from

every other,

of the kind, that

has hitherto appeared,

3.

After having fully explained the Nature

of Equations, and the Methods offinding their

Roots, either infinite exprefjions,

when

it

can
it

be

do?ie,

or in infinite converging feries

to conjider the

remained only

Relation of Equa-

tions involving

two variable quantities, and

to

cf Geometrical Lines
tri?ie

each other

the

Doc-

of

the

Loci
"Tbefe

Equations.

and make

the Conftruction

the

Subject

of of the

Third
*

Part.

Viz..

The

folding of Dh'ifcrs y

and the evolution of

Paitl.

Binomial Surds.

See

59

72. Part II. 127.

Upon

To
Upon
this

the

Re ad e

r.

Plan Mr.

Maclaurin
in the

com-

pqfed a Syjiera g^ Algebra, foon after his being

chofen Profejfor
verftty of

of Mathematkks Edinburgh which he,

;

Uni~

thenceforth,

made nfe of in his ordinary Courfe of Leclitres, and was occajionally improving to the Perfection he intended it

mitted
his

it to the

feu Id have, before he comAnd the beft Copies of Prefs.

to the

Manufcript having been tranfmitted

it

Publifoer,
ejiablijh

was

eafy, by comparing them, to

a correct

a?jd genuine lext*

'There

were, be/ides, feveral detached Papers, fome of which were quite finifed, and wanted only to be iitferted in their proper places. In a few
others, the

Demonfirations werefo

cojicifely

ex-

prefed, and couched in Algebraical characters,

that

it

was
to

necefjary to write

them out at

more
reft.

length,

make them

of

a piece with the

Aid this is

the only liberty the

,

Publifer

has allowed himfelf to take

excepting a few

y inconfiderable additions, that feemed nee efar to

render the Book more compleat within itjelf

and to fave the trouble of confuting who have written on the fame Subjecl.

others

The Rules concerning the Impoffible roots

of Equations, our Author had very fully confidered, as appears from his

Manufcript papers

but as he had no where reduced any thing on

that

To

the

Reader.
what was
it beji

that Subjecl to a better form, than

long ago publifted in the Philofophical Transactions,

N 394, and 408, we thought

of Chap,
11.

to take the fubjlance

Part

II.

from

of thefe Papers a demonjiration of the original ride, furnifoes which prefuppofes only what the Reader has
thence
;

efpecially as the latter

been taught in a preceding Chapter.

The Paper that isjubjoined, on

taken

the

Sums of
is

the Powers of the Roots of an Equation,

from a Letter of
to

the

Author (8 Jul.
the

1743)

the

Right

Honourable
to the

Pari

Stanhope;

communicated

Publifher,

with fome things added by his Lord/hip, which were wanting tojinijk the Demonjiration.

Of
tife is

thefe

Materials, carefully collected

the following
;

and

put in order,

Elementary Treawithin the reach

compofed

which we have chofen rather

is

to

give in a Volume that

of every Student, than in one more pompous, which might be lefs generally ufeful. And we
hope,

from

the pains it has cofl us, its blemifjes

are not ma?iy, nor fuch as a candid intelligent

Reader may not forgive.

The Latin Appendix

is

a proper Sequel, and

a high Improvement, of what had been demon-

frated in
Curve

Part III. concerning the Relation of

lines

and Equations

a Subject which
our

To
quainted with
;

the

Reader.
Orga-

our Author had been early and intimately acwitnefs his Geometria

nica, printed in ij'*g,

twenty- one years of age,

when he was not full and which, though fo

juvenile a work, gained him, at once, that

dijlinguifhed

Rank among Mathematicians^

which

he thencejorth held with great lujlre.

Tet he frankly owns, he was led to many of the

Proportions in
this

Appendix, from a Theorem

of Mr. Cotes, communicated to him, without a?iy demonfir at ion, by the Reverend and

Learned Dr. Smith, College, Cambridge.

Majier of Trinity-

How

he has profited

of that Hint, the Learned will judge : Thus much we can venture to fay, that he
himfelffet fome value upon this Performance ;

having,

we are

told,

employedfeme of the
bequeath
the

lateji

hours he could give

it

to

Juch Studies, in revifng

it

for

the Prefs

-,

to

as his

lafl

Le-

gacy

to the Sciences to

and to

The gentlemen
left

Pub lick. whom Mr. Maclaitrin

are

the care

of

his Papers,

Martin

Folkes,
ety
-,

Efq-,

Prefdent of the Royal SociEfq-,

Andrew Mitchel,
;

Reverend Mr. Hill, Chaplain

Archbifcop of Canterbury

to

and the his Grace the

with

whom

he

had

lived in a mofi intimate friendfiip.

is

And

by their direction this Treatife

publifhed.

CON-

CONTENTS.
PART
Chsp.I.
II.

L
and
Illuftra-

T\EF INITIO NS
*-**
tions

Pag. &

HI.

IV.

V.
VI.
VII.
VIII.

IX.

X.

Of Of Of Of 0/ 0/ 0/ 0/ 0/
Of

Addition
Subtraction
1

Multiplication

12

Divifion
Fractions
/$# Involution of quantities

ij

24 34 42
54
involve
only

Evolution
Proportion

Equations that
quantity

one

unknown

61

XL

the Solution of queftions that pro-

duce fimple equations

6$

XII. Containing fome general Theorems for

the exterminating
ties in

unknown

quanti8

given equations

XIII.

Of

Quadratic equations
Surds

85

XIV. Of

94
127

Supplement to Chap. XIV.

PART

CONTENTS. PART
II.

Of

the Genefis and Refolution of equa-

tions

of

all

degrees

and of the

different

Affedtions of the roots.

Chap.

I.

Of the Genefis and Refolution of equations in general

;

and the number

Pag. 131

of roots an equation of any degree

may have
II.

Of

the Signs

and

Coefficients

of equa-

tions
III.

139
their intermedi-

Of

the Transformation of equations,

and exterminating
ate terms

148
the Roots of equations or more of the roots are
1
1

IV.

Of

finding

when two
V. Of
VI.
VII.

equal to each other

the Limits of equations

62

70

Of

the Refolution

of equations\

all

whofe roots are commenfurate

186
by

Of

the

Refolution

of equations

finding the equations of a lower de-

gree that are their divifors

197

Supplement to Chap. VII. Of the Reduction

of equations by furd divifors
VIII.

213

Of

the

Refolution
rule,

of equations by
others of that

Cardan' s
kind

and

223
Methods by which you may
meral

IX.

Of

the

approximate to the roots of nu-

CONTENTS.
meral
equations

by

their

limits

Pag. 231

Chap.X. Of

the

Method of

Series, by

which

you may approximate

of literal equations

to the roots

244
in

XI. Of

the Rules for finding the number

of impcjfibk roots
tion

an equa-

275
Newton' j
l

XI I

Containing a general DemonJIration of

Sir Ifaac
rule for finding

the [urns of the powers of the roots

of an equation

286

PART
Of the
Chap.
I.

III.

Application of Algebra and Geometry

to

each other.
between the equations

Of the Relation
of curve
thofe curves
',

lines

and

the figure of

in general

297

II.

Of

the Conftruction of quadratic

equations, and of the Properties of

the lines of the fecond order
III.

325
bi-

Of

the Coyftrutlion of cubic

and

quadratic equations

352

APPENDIX.
DeLinearurn Geometricarum Proprietatibus
generalibus,

A TREA-

TREATISE
O
F

ALGEBRA.
PARTI.
CHAP.
Definitions

and

Ilhiftrations.

i.

/WnayBHtf

LGEBR A

is

a general

method

g=

WWwD

=g of computation by certain figns and fymbols which Have been

contrived for this purpofe,

and

found convenient.

It is

called an

Universal
arithmetick,

Arithmetick, and
rules

proceeds by operations and

fimilar to thofe in

common
its

founded upon the fame principles.

ever,
is

This, howufefulnefs or
left

no argument againil
;

evidence

fince arithmetick

is

not to be the

valued

z
valued that

A
it is

Treatise
common, and
clear

of
is

Part

allowed to be
fci-

one of the
ences.

mod
as a

and evident of the

But

number of fymbols
is

are admitit

ed into

this fcience,

being necefTary for giving

its

that extent and generality which

greater!:
is

excellence; the import of thoie fymbols

clearly dated,
arife

to be

that

no obfeurity or error may

from the frequent ufe and complication of

In

them.
2.

Geometry,
triangles

lines

are reprefented

by

a line,

by a

triangle,

and other
;

figures

by a

figure of the

fame kind

but, in

Algebra,
letters

quantities are reprefented

;

of the alphabet

by the fame and various figns have

In

been imagined for reprefenting their affections,

relations,

and dependencies.
are
:

Geometry
in

the

reprefentations

more
the

natural,

Algebra
fir ft

more

arbitrary

the former are like the

expreftion

attempts

towards

of

objects,
\

which was by drawing

latter

their refemblances

the

correfpond more

to the

prefent ufe

of

languages and writing.

Thus

the evidence of

Geometry
vious
\

is

fometimes more fimple and ob-

but the ufe of Algebra more extenfive,

:

and often more ready

thematical
fciences

efpecially fince the

fo

maan
en-

have acquired

vail

extent, and have been applied to fo

quiries.

many

3. In thofe fciences, it is not barely magnitude that is the object of contemplation but
:

there

(Z/v/ieri^ Yta-.

TAB

.1

Chap*
there

i.

ALGEBRA;
many
affections

are

and properties of

quantities, and operations to be performed

upon
In

them, that are necefTarily to be confidered.

eftimating the ratio or proportion of
ties,

quanti-

magnitude only

is

confidered (Efem. 5.
fi-

Def. 3.)

But the nature and properties of

gures depend on the pofition of the lines that bound them, as well as on their magnitude. In
treating of motion, the direction of

motion as
as

well as
that

its

velocity

and the direction of powers

motion,
In
well
as

generate

or deftroy

their forces,

muft be regarded.

Optics,

the pofition, brightnefs, and diftinctnefs of images, are

nefs-,

of no

lefs

importance than their bigis

and the

like

to

be faid of other

fai-

ences.

It is neceffary therefore that

other fym-

bols be admitted into Algebra befide the letters

and numbers which reprefent the magnitude of

quantities.

4. the fign

The

relation

of equality

is

exprefTed

by

thus to exprefs that the quantity

is

reprefented by a
fented by b>

equal to that which

is

repre-

we
is

write a =. b.

But

if

we would

exprefs that a

greater than b 9

we

write

aC~b$

and
lefs

if

than b,

or

5. is capable of being greater or

we would exprefs algebraically that a is we write a __D b. Quantity is what is made up of parts,
lefs.

It is in-

creafed by Addition, and diminifhed

tion
\

by Subtracrations

which are therefore the two primary ope-

4
any quantity
have contrary
a decrement
\

^Treatise
may

of

Part
it is,

L
al-

rations that relate to quantity.

Hence

that

be fuppofed to enter into

gebraic computations two different ways which

effects
,

either as an incre?nent or as
as a quantity to

that

is,

be added,,
fign -ffign

or as a quantity to be fubtracted.
(plus)
is

The
and the

the

mark of

Addition,

(minus) of Subtraction.

Thus

the quantity beis

ing reprefented by a, added, or

-}-

a imports that a
,

to be

reprefents
is

an increment

but

imports that a
a decrement.

to be fubtr acted,

and reprefents

When

feveral fuch quantities are

joined, the figns ferve to

mew

which are to be

added, and which are to be fubtracted.

-j-

a -fand b are both confidered

4- a b denotes the

b denotes the quantity that arifes

Thus when a
and
b.

as increments,

therefore exprelTes the

fum of a and

But

quantity that arifes

when
-

from the quantity a the quantity b is fubtracted y and expreffes the excefs of a above k When
a
is

greater than b 9 then a

,

crement

when a

b,

when

fite,

is lefs

than b y

an o and then a deb

is

itfelf

in-

then a

is

itfelf

crement.
6,

As

addition and fubtraction are oppois

or an increment
is

oppoiite to a decre-

ment, there

an analogous oppofition between

the affections of quantities that are confidered

in the mathematical fciences.
cefs

As between

ex-

and defect

-,

between the value of

effects

or

Chap.
or

i.

ALGEBRA.
,

money due to a man, and money due by him ; a line drawn towards the right, and a line drawn to the left gravity and levity % elevation above the horizon,
it.

and depreffion below

oppofite

When

tw o quantities equal in refpect of

T

magnitude,

but

of thofe

kinds,

are

joined together, and conceived to take place in

the fame fubiect, they deftroy each other's effect,

due
each

to a

amount is nothing. Thus 100/. man and 100/. due by him balance other, and in eftimating his (tock may be
and
their

both neglected. Power is fuftained by an equal power acting on the fame body with a contrary direction,

and neither have

effect.

When

two unequal
ties

quantities

of thofe oppofite quali-

are joined in the

fame

fubject,

the greater
a great-

prevails

by

their difference.
is

And when
leffer

er quantity

taken from a

of

the

fame

kind,
kind.

the remainder

Thus
is

if

we add

becomes of the oppofite the lines AB and BD to-

gether, their

fum
., but

AD
we

r
it

are

C
t-

+
B
+
.

to fubtract

Be
BD
or
is

D
way

from AB,

then

BC

=.

to be taken the contrary

is

towards A, and the remainder

AC

which,
a line

when BD,
forces are

BC
fide

exceeds

AB, becomes
T

on the other

of A.

W hen
2

two powers or
their

to be

added together,

fum acts upon

6
upon
which
other,

A T r e at
the

i s

tf/'

Part

body

but when we are to fubtracl:

one of them from the other, we conceive that

is

to be fubtradted to be a
,

power with an

oppofite direction
it

and

if it

be greater than the,

will prevail

change

of quality

by the difference. Thishowever only takes place

of fuch a nature as to ad-

where the quantity

is

mit of fuch a contrariety or oppofition.

We

know

nothing analogous to
confidered
;

it

in

quantity ab-

ftraclily

and cannot fubtracl a greater

leffer,

quantity of matter from a

or a greater the appliis

quantity of light from a

leffer.

And

cation of this doctrine to any art or fcience

to

be derived from the

fcience.

known

principles of the

quantity that is to be added is like j. wife called a pq/Jtive quantity j and a quantity
to be fubtracled
is

laid to

be negative

they are

equally real, but oppofite to each other, fo as

to take away each other's effect, in any operation,

30,

when

they are equal as to quantity.

and a

am

Thus

o.

and

a are equal as to quantity,

-j-

fuppofe in Algebra that

But though -f a we do not

>

becaufe

to infer equality in this fcierxe, they

mud

not

only be equal

as

to quantity, but of the fame

quality, that in

every operation the one

may
in- all

have the fame

effecl as the other.

A
it

decrement
has

may

be equal to an increment, but

effecl:
,

operations a contrary

motion downwards

Chap.

i.

ALGEBRA.
it

wards may be equal to a motion upwards, and

the depreffion ot a (tar below the horizon

may

be equal to the elevation of a flar above

but thofe pofitions are oppofite, and thediftance

of the
at the
it,

flars is

greater than if one of

as to

them was

horizon fo

have no elevation above

it.

or depreffion below

It is

on account of
is

this contrariety that a negative quantity

faid to

be

lefs

than nothing, becaufe

it

it

is

oppofite to
to
it,

the pofitive, and diminifhes

when joined
effecT:.

whereas the addition of o has no

a negative
is

But

to be confidered no lefs as a real

quantity than the pofitive.

Quantities that have

are underftood to be

no

fign prefixed to

them

pofitive.

8.

The number

prefixed to a letter

is

call-

ed the numeral

coefficient,

and fhews how often

is

the quantity reprefented by the letter

taken.

to

be

Thus 2a imports
is
;

that the quantity re\

prefented by a
is

to be taken twice

$a that

it

to be taken thrice
is

and
is

fo

on.

When

no

number

prefixed,,

unit
1

underftood to be the

coefficient.

Thus

is

the coefficient of a or

off
Quantities are faid to be
are reprefented
like

or fimilar, that

by the fame

letter

qually repeated.

Thus
is

3^ and

a
of

or letters eare like

;

but a and

b,

or a and aa are unlike.

faid

quantity

to confift

as

many

terms as there are parts joined by the figns

+
or

8
or *
;

A
thus a
-f-

Treatise
b confifts of
;

qj

Part

L
is

two terms, and

called a binomial

a-\*b
a

~\- c

oonfifts of three

terms,
called
confifts

and

is

called

trinomial.
:

Thefe are
quantity
or:

compound quantities

a fimple

of one term only,

as -j- a,

or 4- ab^

^
in

abc.

The
places.

other fymbols and definitions necefTary

Algebra fhall

be

explained in

their

proper

C
Of

9.

HA

P.

II.

Addition*
that are

^\ ASE I. To add quantities \^A like and have like figns.

Add

Rule.

together the coefficients* to their

fum

prefix the

common figm and

fubjoin the

common

letter or letters.

EXAMPLES.
To *\~5* Add J-4
to

6b

add

lb

to

a-Y b

add S aJr5^

Sum 4~ 3^

Sum 8^

Sum \a -j- 6b

To

Chap.

2.

ALGEBRA. To 3^ 4x Add $a Sx Sum Sa i2x

To
add
quantities that are like but
figns.

Cafe

II.

have unlike
Rule.
Subtracl
'prefix

the

leffer

coefficient

from
letter

the

greater,

the Jign of the greater to the

remainder', and fubjoin the common

ters.

or

let'

EXAMPLES.
To Add
Sum

4a
-f ya
-j- s a

2^-f-

2<:

To
Add

Sum

+ 4y 4a + 2x y + 5

50
rule
is

6x

57+8
3

2tf

2^
2#

4.x

2a

This

eafily

deduced from, the nature

quantities to be
pofitive together

of pofitive and negative quantities.

If there
are

more than two

firft

added together,

add the

into one fum, and then the negative (by Cafe I.)

Then add

thefe

two fums together (by Cafe

II.)

B 4

EX-

10

Treatise

of

Parti,

EXAMPLE.

la 10a 12a
-\-

8^

Sum Sum

of the pofitive
of the negative

iga
a

-J-

i%a

Sum
Cafe
Rule.
III.

of

all

To
them

add quantities that

all

are unlike.

Set

down

one after another, vnih

their ftgns

and

coefficients prefixed.

EXAMPLES.
To Add

+ +

ia
3*

+ 4#
$
3
i

3<*

Sum

20+
4a

Sa

4*
i

To Add
Sum

+ 4^ + 3c 4* 4? + Z
n
ii

-}-

4^

-4-

3f

4* ~"

4.7

3Z

CHAP,

Chap.

3.

ALGEBRA,

11

CHAP,

10.

III.

Of SUBT R ACTION.
/^I Eneral
rule.
to

Change the Jigns of the

be fubtraffed into their
it

\Jf
quantity

quantity

contrary Jigns,

and then add

it

fo changed to the

from which
lafi

was

to be fubtrafted (by

the rules of the

chapter :) the

fum

arifing by

this addition is the remainder.

For, to fubtract
is

any quantity, either pofitive or negative, fame as to add the oppofite kind.

the

EXAMPLES.
From
Subtract

+ 50 + 34
ia

2a
50

yb

3^4- 4^

Remaind. 50* 30, or 2a

11b
6

From
Subtract

3^ + 5^
-f-

6a

4*

By 4-

Remaind.
It
is

40 yx

10
away a

evident, that to fubtraft or take

is

decrement
crement.

the fame as adding an equal in-

we take away b9 b from a there remains a ; and if we add + b to a b9 the fum is likewife a. In general, the fubtracIf

tion of a negative quantity

is

equivalent to add-

ing

its

pofitive value.

C HfA

P.

\Z

Treatise

of

Parti.

CHAP.

IV.

Of Multiplication.

n.TN

Multiplication the General rule for

is,

JL the figns
the factors are like

That when
both
\

the figns

(i. e.

+,

or both

of
,)

the fign of the

produB

is -j-

but when the figns

of the faftors are

is

unlike^ the fign of the produft

Cafe

I.

When
-J- n,

any pofitive quantity,

+
+

a,

is

multiplied

by any pofitive number, the meaning is, That a is to

as
;

be taken
units in n

many

times as there are

is

and the product

evi-

dently na.

Cafe

II.

When

is

multiplied

by

n,

then

is

to be taken as often as there are

units in n>
na.

and the product muft be

Cafe III. Multiplication by a pofitive number implies a repeated addition

:

but

multire-

plication

by a negative implies a

peated fubtraction.
to be multiplied by

And when

-4-

is

n>

the mean-

ing

is,

That

is

to be fubtracted as

often as there are units in

n:

there-

fore

Tab

it

Chap. 4

ALGEBRA.
the

13
being

fore

product

is

negative,

na.

Cafe IV.

there are units in n

fubtract
is

a a then ^
When
is

is

to be multipled

by

to be fubtracted as often as
\

but (by 10.) to

equivalent to adding
the product
is

+ a, confequently
The
By we
lid and

+ na.

IVth

cafes

may be
**

illuftrated in

the following manner.

the definitions,

a by #, the product mud multiply -f a vanim or be o, becaufe the factor a a is o.

if

+ a

=
is

>

therefore,

The
Cafe

firft

term of the product

+
-\-

na (by na

I.)

Therefore the fecond term of the pro-

duct muft be

na^

which deftroys

that the whole

product muft be

o.

Therefore

a
if

na

na
gives

fo

multiplied by

+n

-*- na.

In like manner,

we multiply

+aa

a, the firft

term of the product being

by
na>

the latter term of the product muft be -{- nay

becaufe the two together muft deftroy each other,

or their amount be o, fince one of the factors

n muft give
gny kind

(viz.

a-

a)

is o.

Therefore

a multiplied by
is

-f- na.

In this general doctrine the multiplicator

always confidered as a number.

quantity of
:

may

be multiplied by a number

but

a pound

H
a pound
is

Treatise ef
line

Part

not to be multiplied by a pound,

or a debt by a debt, or a

by a

line.

We
is

mall afterwards confider the analogy there

betwixt rectangles in

Geometry and a product

of two factors. 12. If the quantities to be multiplied are

quantities, find the ftgn of the produtl by
;

Jimpk

the. loft rule

efficients,

after it place the produtl of the co-

and then

fet

down

aft the letters after

one another as in one word,

EXAMPLES,
Mult.

By
Prod..

+a +b

* la

+ Ab

6x
5a

+ ab

Sab
Sx
4-a

%cax

Mult.

By
Prod.

+32 ax

igaabc

13. To multiply compound quantities, you muft multiply every part of the multiplicand by
all the

parts of the multiplier taken one after an-

ether,

and then
that

collecl

all the producls into

am

fum

fum

fhall be the produtl required,

X**

Chap.

4.

ALGEBRA.
EXAMPLES.
a

Mult.

b
b

2a

%b
Sh

By
Prod,

a+
aa-{-

4^+
Saa

ab

I2ab
-\-10ab--15bb

ab+bb
'

\
$aa

Sum
Mult.

aa+2ab + bb
2a

xx

2ab

i$bb

4b
Sab
+
Sabi6bb

ax

By
Prod.

2a-\-4b

x+a
xxx ###

\aa

I
400

1
xxx

+ ^## aax
. .

Sum

..

o ~i6bb

aax

Mult.

0*-[~^ -f*^
C aaa
2,

By
-\-

aab
tf tf

<3tf#
.

abb bbb
.

-f-

^^
O

Su m

JM

14, Products that arife from the multiplication of two, three, or more quantities, as
abc 9 are faid to be of two, three, or
menfions
\

more

di*

and thofe quantities

are called faolors

or

roots*

If

i6
If
all

^
of
a.

Treatise
,

of

Part

the factors are equal, then thefe proas aa or aaa are powPowers are expreffed fometimes by

duels are called powers

ers

placing above the root to the right-hand a figure

expreiTing the

number of

factors that

produce

them.

Thus,
a
UJ

"ift^ Power of the

aa

2d
)>

root a\ and
is

aaa
aaaa
aaaaa
^

fhortly<

4th
?9

exprefifed

:3*.

thus,

15. Thefe figures which

their indices or exponents

ex-prefs the

num-

ber of factors that produce powers are called

;

thus 2

is

the index of

-.

And powers
their

of the fame root are multiplied

by adding

exponents.
1

Thus
ufeful

az

a1

=a

a^

Xa

= a\

=
it

#4.
is

16. Sometimes

not actually to

multiply

compound

quantities,

but to

fet

them

down

with the fign of multiplication (x) between them, drawing a line over each of the

compound
prefies

factors.

Thus

a-\- b

b ex-

the

product of a

b multiplied by

b.

CHAP.

Chap,

5.

ALGEBRA.

ty

CHAR
17.

V.

Of Division.

ryiHE
JL
\

fame rule

for

the figns

is

to

be obferved in Divifion as in Multithat

is,

plication

If the figns of the dividend

the fign of the quotient muft

and
be

divifor are like,

v

if they

tient muft

be

are unlike\ the fign of the quo.

This

will

be eafily deduced

from the

rule in Multiplication,, if

you confider

that the quotient

muft be fuch a quantity as

multiplied

by the divifor mail give the divi-

dend.

18.

The General

rule in Divifion

is,

to place

the dividend above a fmall line,

and the

divifor

under

it,

expunging any

letters that

may

be

found

in all the quantities

of the dividend and divifor*

of
all the

and dividing the

loab

coefficients

terms by

any common meafure.

Thus when you divide

all
all

i$ac by load, expunging a out of

the coefficients

the terms, and dividing

the quotient
is

by

5,

~Jf
4
>.b)ab

and

b a + bb( +

2ab) $oax

way ( 5 *^ 97

4^)

i8

Treatise
-j-

of
.

Psrrtl
.

^aa) Sab

6ac

And 2^) 5^^ f

19. Powers of the fame root are divided by
fubtratting their exponents as they are multiplied

by adding them.
the quotient
is

Thus
z
*,

if
? .

you divide a 5 by a z >

a^~ or #

And
7
!

6
/>

divided by
l

64 or b z gives Z>
5

and # divided by a z b

gives * for the quotient.

20.

If the

quantity

to
its

be divided

is

com-

pound, then you mufi range

parts according to
as in the

the dimenfions of fome one of

its letters,

following example.

In the dividend a z -{- 2ab

+b

z
,

they are ranged according to the dimenfions of

a,
1 the quantity a where a

is

of two dimenit

fions being placed

firft,
% ,

lab where

is

of one
all,

dimenfion next, and b

being placed
laft.

where a

is

not at

The

divifcr mufi be ranged ac-

cording to the dimenfions of the

fame

letters

-,

then

you are

to divide the firft

term of the dividend by

the firft term of the divifor, and to fet

quotient,

down

the

which, in this example?

is

then mul-

tiply this quotient

by the whole divifor,

and fubthe re-

trail the producl

from

the dividend,

and

mainder fhall give a new dividend, which in this

example
is

ab-\-b z .

Chap.

5.

ALGEBRA,
+ a? +
a%
lab
&b

39

+ b)

+b

+h

ab

+b ab + b

z
z

Divide the
firft

fir ft

term of this new dividend by the

b) with

term of the divifor and fet down the quotient^

this

(which in

example

is

its

proper Jign.

Then

multiply the whole divifor by this part of the

quotient^

dividend
is

and fubtratl the produtl from the new and if there is no remainder, the divifion
:

finifhed

If there

is

a remainder,
till

you

are to

proceed after the fame manner

is left
,

no remainder

or

till it

appear that there will be always

fome remainder.

Some Examples

will illuftrate this operation*

EXAMPLE
a

I.

+ b)a
a

b
-\*

(a

b
l

ab

ab

4b

EL

20

^Treatise

of

Parti

EXAMPLE aaa gaab + gabb a

b)

II.

aaa aab 2aab + gabb bbb labb abb bbb

iaab-\-

bbb (aa

lab \ bh

abb

bbb

EXAMPLE aaa a + ab
b)

III.

bbb {aa

-\-

bfr

aaa aab

aab

aab

bbb abb
abb

bbb

abb bib

EX-

chap.

5.

Algebra.

21

EXAMPLE
j# ~~
6)

IV.
-{-

6aaaa
6aaaa

96
;

{jiaaa

+ \cia

Sa

+ 16

laaa

96 i^aa iiaaa
naaa
rL\aa

i^aa

96 48^
48*2
4.8a

96

96

may

21. It often happens that the operation be continued without end, and then you
infinite

have an

Series

for the quotient

and by

wnparing
find

the firft three of four terms, the terms obferve

:

you may

what law

by which means ^

without any more divifion^ you may continue the

quotient as
1

by

far as you pie ofie.

ai

Thus,
which

in dividing

you find the quotient

as

to be

+d
be

+ aa-\- aaa + aaaa + &c

continued as far

Series can

you

pleafe

by adding the

powers of

a.

Q2

The

22

A
The
I

Treatise
Operation
( i

of

Parti.

is

thus
*
"

a)

+ a + aa + aaa & c
a

-^-a-aa

JLaa'

aaa
-\-aaa

if aaa

aaaa
+
aaaa^ &c.

Another Example.
a

+ x)

aa

+ xx (a x-{ aa + ax
ax-\-xx
ax

2xx

2x %

~r

+ ^

^a*

4"

pj &

xx
-\-2XX
.

A-2XX

2x

2* 3
a

2x %

2X A

In

(Z/i/ten

Jia

22.

TAB.UT
&ty..t 0.71.3

c ___

&ty zo

&&.*3

Chap.

5.

ALGEBRA.
laft
,

23

In this

example the

figns are alternately

+
the

and
firft

the coefficient

is

conftantly 2, after

two terms, and the letters are the powers of x and a ; fo that the quotient may be continued as far as you pleafe without any more
divifion.

But

in Divifion, .after

you come

to a remain-

der of one term, as ixx in the

it
it,
is

laft

example,

commonly
the

fet

down

with the divifor under

after

other

terms,

and

thefe

together
in the
.

give the quotient.

laft

Thus, the quotient

example r

is

found to be a
b

lx x
-\
;

bb

+ ab divided by
1

a-\-x

And

gives for the quotient

Note>

The

fign -r placed

between any two

quantities, expreffes the quotient of the

former

divided by the
the quotient of a

latter.
-\-

Thus

a-\-b

-7-

b divided

by a

is

x.

CHAP.

24

A
:

Treatise

of

part

CHAP.

VI.

Of Fractions.
22.

fN
it.

the lafl Chapter

it

was

laid that the

it quotient of any quantity ^ divided by

b
is

expreiTed by placing a above a fmall line and

I under

ihus 3
;

-r.

Thefe quotients are

aifo

called Fratllons
I

and the dividend or quantity

line
is

placed above the

called

the Numerator

of the fraflion,
2 exprefles

and the diviior or quantity

is

placed under the line

called the Denominator,

Thus
and

the quotient of 2 divided

by

2 is

the numerator and 3 the denominator of

the fraction. 23. If the numerator of a fraction

io
is

equal

the denominator, then the fraction

is

equal to

unity.

Thus
is

and

are equal io unit.

If the

numerator

greater than the denominator, then

the fraclion is greater than unit.

cafes,

In both thefe

the fraction
is

is

called

improper.

But

if

the

numerator

lefs

than the denominator, then

is

the fraclion

is lefs

than unit, and

called proper.
$

Thus
rr are

is

an improper fraction

but

is

and
that

proper fractions,

mixt quantity

whereof

Chap.

6.

ALGEBRA.
is

25

whereof one part

a fraclion.

an integer and the other

5-^,

As

3-^

and

and a

~.

PROBLEM
24. 7*0 reduce a

I.

mixt

quantity to

an impro-

per Fraction.
Rule.
Multiply the part that
is

an integer by the
\

denominator of the fractional part

and

to the

producl

add

the numerator

under their

fum

place the former denominator.

Thus 2~ reduced
gives

to an

improper fraction,
3

-s

a -J- -:

and a

x4

PROBLEM
mixt Quantity.
Rule.

II.

25. To reduce an improper fraclion

to

Divide the numerator of the fraclion by

the denominator,
integral part
\

and the

quotient fhall give the

the remainder fet over the deno-

minator fhall be the fractional part.

1-T-1

12

ab-\-a x

cr

ax4-2xx

x%
a-\-x

aaA-xx
a-~-x

2xx

'

ax

C 4

PRO-

26

Treatise

5/

Part

PROBLEM
26. To
tions to fractions of equal value

III.

reduce fractions of different denomina-

that Jhall have

the fame denominator.

Rule.

Multiply each

nuttier-ator,

feparately taken*

into all the denominators but its

own, and the

'products jhall give the

new

numerators.

Then

multiply all the denominators into one another,

and
tor.

the produft Jhall give the

common denomina-

Thus

the fractions

, ,

are refpectively

equal to thefe fractions 7. ^ bed

-7-7,

-77,

which

bed

bed

have the fame denominator

1
3

3 4
,

bed.

And

the fractions

are

^r^:.^i equal refpe&ively _..i

4 ^ ^r- ^-, to thefe

45
60'

4
60'

PROBLEM
27.
Rule.
7*0

IV.
Fractions*

add

fli

subtract

to a common denominator* and add or fubtracl the numerators, the fum

Reduce them

er difference fet over the common denominator*

js the

fum

cr remainder required.

Thus

Chap.
_,
a

6.

ALGEBRA.
+ + .bat =
,

27
bee
a -f- </ \

Thus

ade

-r

-r

ti bde

L
d
s

ad

hc

4"*"

bd 3

'

A ""
4
9

+9 _ ~~
12
1
a

7
12

2
"3
...

12
5

8
m

""

4 __
A*

54
J3

l6

20

"" 12

*j. 20

2 X ~~~ 2X

PROBLEM
28.
Rule.
to

V.
Fractions.

To

multiply

Multiply their numerators one into another

obtain the numerator of the produfil,

and

their

denominators

multiplied into

one another

Jhall give the denominator of the produtt.

a
e

Thus

~T~ X
it

+b

T x 7 = ^; a*b* ab
d

ac

2 8 4 -x-sr

>

and

cd

'

If a mixt quantity

is

to be multiplied,

firft

reduce

to the form of a
is

fraction (by Prob. 1.)

And
tion,

if

an integer

to be multiplied
it

by a

frac-

you may reduce

to the
it.

form of a

frac-

tion

by placing unit under

EXAMPLES.
*

51

292

28

A
18 3

Treatise
bx + a

of

Part!

=
tfb

== 6
tf#

bx

==

ax

PROBLEM
29.

VL

fo divide Fractions.

Rule.

Multiply the numerator of the dividend by

of the divifor,
their product

the denominator

fhall give the numerator of the quotient.

Then

multiply the denominator of the dividend by the

numerator of the divifor, and their pro duel Jh all

give the denominator.

Thus

4\ 2 /io 5 y 3 li2'
a

3\
b

5/35

C \

a (ad

+b a- b)

jJSXu" JJJVcb
,'
\

2ab -f
a"
-\-

ab

30. Thefe
ftrated
i

lafl

four Rules are eafily

demon-

from the

definition of a fraction.

It is

obvious that the fractions

b
9
,

are refpectively equal to

-^,

-^r,

^7-7,

fince if

you divide adf by bdf the quotient

fame
~dbf
as

will

be the

and cbf divided by gives the fame quotient as c divided by d ;

of a divided by b
-,

and ehd divided by fbd the fame quotient vided by /.

2,

as e di-

Fractions reduced to the fame denomina-

tor are added

by adding

their

numerators and

fob.

Chap.

6.

ALGEBRA.
the

&g
I

fubfcribing
t. b

common
For
be a
o

denominator.
call

fay

4.
'

L
b

^-_f.

4 = o
r

wh and
fl>

&*='% and

it

will

#?,

and

^
is

= + and m + = y- that + = ^4~~* After the fame manner, b b b

-f-

/;

f,

>

a
~b

a ~~
fay
'

~ b
3.
I

- "' ~ "' bd y X 7 (=XJ=^i ~ j

^

forte

= ^,
that

dn
1S
'

c\
c

and bdmn
ac

*f,

and w

-7-^ ;

7x7
4.
I fay

~~

w'
divided by

-7-

or

nd
CO

gives

and
is,

for

mb

=
c

a,

and

^i

-=.

ad\

=
;

c,

#J

=
'

therefore

-7-7
Wtf

that

m
n

ad
cb'

PROBLEM
31. To
two numbers
find the greatefi
\

VII.

common Meafure of

that

is,

the greatefi

number

that

can divide them both without a remainder.

Rule.
lejfer,

Firft

divide

the greater
is

number by the
lejfer'

and if there
is

no remainder the

number

the greatift

common

divifor required.

30
If
by
laft

A
there
it
;

Treatise

of

Parti.
laft

is

a remainder, divide ycur

divifor

and thus pre ceed coniinuaVy dividing the

till

divifor by its remainder,

left,

there

is

no reis

mainder
greateft

and then the

laft

divifor

the

common Meafure

required.

Thus

the greateft

63 is 9 ; and the 256 and 48 is 16.

45) 6 3( J

common meafure of 45 and greateft common meafure of

48)256(5 240
16)48(3
48
.

45

18)45(2
S6
9) 18 (2 18

32.

Much
,

after the

fame manner the greateft

is

common
covered

meafure of algebraic quantities

only the remainders that

dif-

arife in

the

operation are to be divided by their fimple divifors*

and

the quantities are always to be ranged accord-

ing to the dimenjions of the

fame

letter.

Thus
of a
z

to find the greateft

and a
b
x

lab + b
b*

common
z ;

meafure

a a

lab + %ab +

b* (1

2#* Remainder,

which

60z/t&?v

Jtefr

$1

Chap.
which

ALGEBRA, reduced divided by lb ^.b)a +b a

6.
is
z

31
to

(a

o
Therefore a
fure required.

o
the greater!:

is

common mea-

The ground of
mainder
dividend
(if
-,

this operation is,

That any

quantity that meafures the divifor and the rethere

is

any) muft alfo meafuie the

is

becaufe the dividend

equal to the

fum of

the divifor multiplied into the quotient,

*.

and of the remainder added together

in the laft

a1

example, a

Thus
;

b meafures the divifor

and the remainder

lab + 2b

muft
to

therefore like wife meafure their

.

fum a z

it

2ab -f b z

You muft
letter,

obferve in this operation

make

that the dividend which has the higheft

powers of the

according to which the

quantities are ranged.

PROBLEM
33.
terms*

VIII.
to
its

To

reduce avy

Fraction

lowefi

Rule.

Find the

greatefl

common meafure of the

;

numerator and denominator

divide them by that

common meafure and

place the quotients in their

room, and you Jhall have a fraclion equivalent

* See

Chap. XIV,
ti

32
to

Treatise

of
in

Part
the

the given frafiion

exprejfed

leaft

terms.

Thus

lia
h
z

7Sah
)

==

5#

5 6aa

1 5 6ah

-J

2$bc) I2$bcx

'

$*] 2aa -,

57 2ah

lb
_

'

a?

2ab
'

-j-

az

a3

ba

a*
a
s

b*
7

jrqr^T+T
When
is

~~

<?

+b

_
'~~

a*

+#
'

a*b

unit

is

the greateft

common
Thus

meafure

of the numbers and quantities, then the fraction

already in
its

loweft terms.

- can-

not be reduced lower.

And
fure
is

numbers whofe
unit, are faid to

greateft

common mea-

be prime to one another.

34. If

it is

required to reduce a given fracit

tion to a fraction equal to

that fhall have

given denominator, you muft multiply the numerator by the given denominator, and divide the

product by the former denominator $ the quotient

fet over the given denominator
quired.
is

the fraffion reit

Thus

it

being given, and

being re-

quired to reduce

to an equal fraction whofe dec

it
;

nominator

fhail

be

find the quotient of ac

divided by #, and

mail be the numerator of

the fraction required,

ft

Chap.

6.

ALGEBRA.
is

33

If a Vulgar fraction

to be reduced to a

decimal (that
is

is,

a fraction
its

whofe denominator

10,

or any of

powers) annex as many

cyphers as you pleafe to the numerator*

divide
it

and then

by the denominator*
to the

the

quotient Jhall

give a decimal equal

pofed.

Vulgar fraction pro-

Thus

=
=

.66666* &c.

-2.

.6*

35.

.2857142, &c.

Thefe

fractions

are
*

added and fubonly

care

tracted like

whole numbers

muft

be taken to
as units

fet fimilar places above one another*

above

units,

C5V.

They

are multiplied
;

and tenths above tenths, and divided as inte-

ger numbers

only there muft be as many decimal

places in the pro duel as in both the multiplicand

and

multiplier

and

in

the quotient as

many as

there are in the dividend more than in the divifor.

And

in divifion the quotient

may be
pleafe,

continued

to any degree of exactnefs

you

by add-

ing cyphers to the dividend.

thefe operations
is

The ground of
and divid-

eafily

underftood from the ge-

neral rules for adding, multiplying,

ing fractions.

CHAP.

34

-^

Treatise

of

PartL

CHAP.
Of the Involution

VII.

of Quantities.

36.rTTMIE
JL

produces arifing from the conmultiplication of the

tinual

fame

quantity were called (in Chap* 4.)

the powers

z of that quantity. Thus #, a y a\ <**, &c. are z the powers of a ; and ah* a b* 9 a b\ a^b\ Sec.
3

are the powers of ab.

In the fame Chapter,

the rule for the multiplication of powers of the

fame quantity

is

to

" Add
;

the exponents and

make Thus

their

fum
a?

the exponent of the product."

a*

X
5.

a9

and a*b*

a bz
for

= ab
9
-'

s
.

In Chap.

you have the

rule

dividing

To powers of the fame quantity, which is, exponents and make the difference fubtract the
the exponent of the quotient."

Thus

a = a*

6 """

a*

and

a5P

~r a*b

7=

37. If you divide a

leffer

the exponent of the quotient muft,

power by a greater by this Rule,

',

be negative.
<**

Thus -^
and hence

=
a
1
A

a* 6
is

a \

But by a %

-5

1 ~i

- ;

exprefled alfo

with a negative exponent.

It

(Z/i/ie<ns

4m&

34

TAB

&#.**.?

Chap*
It
is

7.

ALGEBRA.
obvious that

35
a
1

alfo

= #$

but

i,

and
1

therefore a

==

1.

After

the fame

manner

$ = =

a
###
1,

aa
tf"~
3

a
;

'

fo that the quantities ,

&c.
z
9

may

be exprefTed thus, a\ a* 9 a~~\

,

a~\ a~ 4
;

&c.

Thofe

are called

the

negative powers of a

which have negative ex-

ponents

but they are at the fame time pofitive

powers of

or

3$. Negative powers

multiplied by

(as well as pofitive) are

adding*

and divided by fubtrafting

the product of (or

their exponents.

Thus

a~ %

(or
5

A
(or

multiplied by

a~
x
%

-M

is

a~ z "" 3
==

= a

4)
;

alfotf-

a*

a~ 6 + *
1.

a~ z
in

(or

and

a x
3

= a =

And,

gene-

ral,

#v pofitive power of a multiplied by a negative

power of a of an equal exponent gives unit for the product \ for the pofitive and negative deftroy
each other, and the product gives a
equal to unit.
,

which

is

Like-

3.6

^Treatise
-

of
a~~ 3 =.

Part

Like wife
a

==

<s~~

+2

as

and
r

~^i ==
fc=

# +
~
;

<s

3 .

But

alio,

r:

=
^

=a

therefore

tij

^3

And,

in general,

" any

quantity placed in the denominator of a

fraction

may be
of
its

tranfpofed to the numerator,

if the fign

exponent be changed."

Thus

and

3.

39. The quantity a* expreiTes any power the exponent (m) being undeof a in general
*,

termined; and

a~ m

expreiTes

a
:

or a negative

power of a of an equal exponent

and a m
B

x a~ m

product. g expreiTes a m + n is the a m x a" product of the powers am and #", and s* * is
7;/

(3

w =5 # '==

1 is their
;

any other power of a

their quotient.

40.

To

raife

any fimple quantity to

is

its

fe-

cond, third, or fourth power,

to

add
to

its

ex-

ponent twice,
therefore

thrice,

or

four times

itfelf
is

the fecond
its

power of any quantity

and,
in

had by doubling
trebling
its

exponent, and the third by

5

exponent

general,
is

the

power

exprejfed by

of any quantity

had by

multiplying the exponent by

as

is

obvious from
the fecond
, 1

the multiplication of powers.

Thus
*
1

power or fquare of a

is

az

its

third

power

Chap.

7.

ALGEBRA.
is

37
is

power or cube
of a
is
1

ai*
2 .

= a>\
a** is

and the mth power

^3x4

^x = ^
8
<2
;

Alfo, the fquare _of a*

^X4abc
l

the cube of
is is

lz
\

and

the /th power of a*

is

#4X,

The
9

iquare of

a b

%
9

the cube

a b*c

the sth power"

41.
is

The

raifing of quantities to
,

any power
is

called Involution

and

<2/{y

fimple quantity

involved by multiplying the exponent by that of the

power
The

required^ as in the
coefficient

preceding Examples.
the

muft alfo be ratfed to

power by
itfelf,

a continual multiplication of itfelf

as unit is contained in the

fame by

as often

expo-

nent of the power required.

Thus

the cube of

%ab

is

X &b
it is

=z 2ja*bK

As
volved

to the Signs, V/hen the quantity to be inis pojitive,

obvious that all

its

powers

muft be pofiiive.
involved
is

And when

the quantity to be

negative, yet all its

powers whofe ext

ponents are even numbers muft be pcjitive^ for

any

number of multiplications of a negative, number is even, gives a pofitivej fince

=z 4. , therefore

= -f
its

and

X x X X = + X -f xX XXX =
then only can be negative
is

if the

+*+

+ ==+.
when
an odd

The power
exponent

number, though the

quantity to be involved be negative.

ers

of

a are

The pow-

a,

+ *S D 2

*'%

+a

*i

~~ a

*<>

&f<

38
&c.
are
J
3

A
pofitive
3?
,

Treatise

of

Part

I.

Thofe whofe exponents

5s &SV- are negative.

are 2, 4, 6,

&V.
arc

but thofe whole exponents

42.
is

The

involution of compound quantities

more

difficult operation.

The powers
follows.

of

any binomial a-\- b are found by a continual multiplication of

it

by

itfeif as

X
ft
ft

X
*
ft
ft

X
^
ft ft ft

X
*
ft

X
%
ft
ft
53i

ft

ft

&

ft

ft

+ + + ++ ++ ++
Ln
c>
<J

4-.+
OO
ft

4ft
(j

+-
to
ft

<5>Oi
ft
+.

4ft

++ ++
i?> to
ft

ft
t

ft

ft

ft--** ^ ft
ft-

ss-Oo ^
ft-

ft

+ + + + ++ + ++ + ++ + H M
t-i

<5-

ft-

^-

ft-

^
+

O
o

ft- ft-

ft ft-

ft-

-f"
fti

Cn
ft

<-M
ft

O Si

O
ft
'

4^

O ^
<5~

ft-

C-O
ft <s
ft

00
ft

00
ft ft-

{O
ft

ft

ft--

ft- ft-

ft*

55-

<- ft-

<a- ft-

II
rti

+
10
Si

+
ON-^
ft

++
ft

O O
Ki
Si

O
ft^

ft

ftS

<

<5>

4- 00 ft ft 0- <5-

ft"

II

CO ^ c

ft- ft-

+ t+ + ++ + t
ft*

O c
i-(

Cn
ft

O
ft

Oi
ft

Cn
ft

>
ft

1
to

iJ

^j-

ft-

+ ++ + O
<-a
ft
VI

ft- ft-

ft <5- ?s4> +

li

0Go

-a
f6

ft-

J5*

5s

^4 ^.
II

ft

p p

H3

*'

3 O
O

++
0k

"

Oi
pr
73

II

43'

(X/z/tevt

fuz

gg

Tab

rvx

^%

35

iT'ta

Chap.

7.

are required, 43. If the they will be found the fame as the preceding,

ALGEBRA. powers of a b
number of

39

only the terms in which the exponent of b

is

an odd number will be found negative

caufe

-,

" beof a
z

an odd

multiplications

negative produces a negative."

of a %ab
z

b
b
l

Thus

will
:

be found to

be a 1

%a b

the cube
-j-

where the ?d and 4th terms are neIn

general,

gative, the exponent of b being an

in thefe terms.

any power of a

odd number " The terms of

and negative

b are pofitive

by turns."
44.
It is to

be cbferved, That

"

in the firft

term of any power of a

b,

the quantity a

has the exponent of the power required, that

in the following terms, the

exponents of a de-

creafe

gradually by the fame difference (viz,

unit;

and

that

in

the

lafc

terms
in

it

is

never

found.
order
,

The
it

powers of b are

the contrary

is

not found in the

the

firfl
is

term, but
unit,
in
its

its

exponent
third

in

fecond term
is

the

term

its

exponent
till

and thus
term
it

ex-

ponent

increafes,

in the laft

becomes

equal to the exponent of the power required."

As
the

the exponents of a thus decreafe, and at

increafe,
'*

fame time thofe of b

their

the

fum
is

of

exponents

is

always the fame, and

equal to the exponent of the power required."

Thus

in the
j

6th power of a
%

b,

viz
5

~f-

6a' b 4-

$cib

-f

zoa'fr

+ kg**** + 6ab
3

4- b\
the

40
the
6,
in

A
exponents
5,

Treatise
of
i,

of
in
this

Parti,
order,

decreafe

4,

3,

2,

o;

and thofe of b increafe

o,
i,

the

contrary

order,
their

2,

2,

4,

5,

6,
is

And

the
6.

fum of

exponents

in

any term

always

45. the

To
you

find

the coefficient of any term,

coefficient
,

of the

preceding

term

being

known

are to

divide the coefficient of

the preceding term by the exponent of b in the

given term, and to multiply the quotient by the

exponent of a
unit."

in the

fame term, increafed by

the
coefficients
-j- b,

Thus

to

find

of the
find the

terms of the 6th power of a

you
5

terms are
a
6
,

a>b,

ab\ a b\ a''b\ ab\ b

the coefficient of the

and you
unit
;

know

firft

term

is

therefore, according to the rule, the coeffi-

cient of the

2d term

will

be

x 1

5 J

-,

that of the 3d term will be that of the 4th term wil 1 be

X4 + i = 3X5 = 15.
*5 x 3+
will
1

== 5 * 4

= 2o
i
%

and thofe of the following

be 15, 6,

agreeable to the preceding Table.

46. In general, if <^ -f- ^ Is to be raifed to any power m, the terms, without their coeffi~ m m cients, will be, a > a m J b, a m z b\ a m ~ib\ a *b*

a-b\ &c. continued

till

the exponent of b

becomes equal

to

m*
The

Chap.

7,

ALGEBRA,
rule,

41

The

coefficients

of the refpedtive terms, ac-

cording to the lad

m i,m,mx 2 m1 m

1
5

m wX
m

&c. more than

T

4 3 4 5 continued until you have one coefficient

there are units in m.
therefore

7 -,

m X

232
will be
1

X
X

m -i

mX
?

m 2

mA.
-S

It follows

by
7
I

thefe lad rules, that

a + b == am

+
,

fncr'-^b -f- ^2

772

tf

~2

+ ^
X
m
3

X
2

23
X
772
*?

a m -3b z
,

+mX
,
.

'2
.
.

X
4

^ a m *b 4
,

+,

l3c.

which

is

the

general Theorem for raifing a quantity confiding

of two terms to any power m.

47.
If a quantity
is

confiding of three, or
c;

more terms
dinguifh
it

to be involved,

you may
it

di-

into

two
it

parts,

confidering

as a

binomial, and

ceding rules

-,

any power by the preand then by the fame rules you

raife

to

may fubditute inftead of the powers of 9> compound parts their values.
Thus a
2c

thefe

+ c .'== a + b a+b + a + lab

-

-\-

= a\ b
1
/3

-}-

c ,-==i

-f" 2tf*

And
-f- 3c
2

-\-

-\-

z=i

-\-

-[2

%c

X
3

+b
5 "4" ^

3**'

+ ^ -r
6

-j-

'

=
3**'

^ 3 ~r 3^ ^ ~r

3^*

4~

3^
4

+ 3^ + ^In

42
In thefe
as

^Treatise of Examples, a + b +
c

Parti.
is

confidered

compofed of the compound part a + b and the fimple part c ; and then the powers of b are formed by the preceding rules, and a

fubitituted for a

+b

and a -f b '

CHAR
43. f"jgp

VIII.

Of Evolution,

HE

reverfe

of Involution,

or the

JL
is

refolving of powers into their roots,

called Evolution,

The

roots of fingle quan-

tities

are eafily extracted hy dividing their expo8

nents by the number thai denominates the root required.

Thus

the fquare root of a

a*b*c
2

is
z

a*

a4

and the fquare root of

cube root of a*b l
root of x?y^z
rule
is
lz is

is

a b+c.

The
of this

is

a 3 b* == *i
4
.

and the cube

x^z

The ground

obvious from the rule for Involution.

root are

The powers of any

plying
its

found by multi-

exponent by the index that denomi,

nates the
is

power

and therefore, when any power

given, the root muft be found by dividing

the exponent of the given

power by

the

number
is

that denominates the kind

of root that

re-

quired.

49-

d/i/ien.fui. 4-$

Tat*

.tic

&&-S&

Chap.

8.
It

ALGEBRA.
"
either

43

49.

appears from what was faid of Invo-

lution, that

any power that has a pofitive fign

may have
root
is

pofitive or negative root^ if the

denominated by any even

z

number"

the fquare root of

becaufe -\~ a

~J-

a or

+ a may be + a or a X a gives + *P
fign,

a
"

Thus
9

for the product.

But
root of

if a
it

power have a negative

no

denominated by an even number can be

is

ajfigned" fince there

plied into itfelf an even

no quantity that multinumber of times can

give a negative product.

of a cannot be affigned, and is what we an " impoffible or imaginary quantity.*'

Thus

the fquare root

call

But
the

if

the

root to be extracted

is

denominum6
3

nated by an odd number, then Jhall the fign of

rod

be the

fame as
is

the fign of the given

ber whofe root

of

a
"

a
z

is

required.

Thus

the cube root

is

a,

and the cube root of a b

b.

% 50. If the root required is

number

that denominates the

a divifor of the exponent

of
the

the given power, then fhall the root be only a

lower power of the fame quantity"

11

As

cube root of a

is a 4 ,

the

number

that de-

nominates the cube root being a divifor of 12.

But
fort

if

the
is

number
required

that denominates
is

what

of root

not a divifor of the

exponent of the given power,

re quired fhall

"

then the root

its

have a fraclion for

exponent."

Thus

44
Thus
of

-A

Treatise
is

of

Part

I.

the fquare root of a*

V^

the cube root

itfelf is

a* is

a\
cc

and the fquare of a

Imperfeft powers or

a 7-.

Thefe powers that have

are called

fractional exponents

y#r^

,"

and are

otherwife exprerled by placing the given

within the radical

the radical

%n
the
is

V~%

power and placing above

that denominates
2.

%n

number

what kind of root

required.

Thus a* =. V a*
In

a?

VV;

and ^*
is

=
4.

vV*.

numbers the

fquare root of 2

expreffed by

2,

and the

7 cube root of 4 by v

51.
<c

Thefe imperfect powers or furds are

and
divided,
as

multiplied

other powers,
9

by

adding and fuhtrafting

their

exponents."
3

Thus
L

a^Xa

7
-

^z

a3

Xa*~a
a

=: a

=z

vV

and -
a?
are

=#

== #

2
.

They
the fame

involved likewife and evolved after

as perfect

manner

powers.
;

Thus

the
*

fquare of a

7
-

is

#~

2-X
~

= a*
#
4
.

i-

the cube of a 5
is
z ^ 3 x =z

is

=za

5
.

The

fquare root of a 3
is

a'

the cube root of a*

But we

(hall

have

occafion to treat 52.

more

fully of

Surds hereafter.

The

fquare

root of. any

compound
this

quantity, as a %

lah

is

difcovered after

Chap.
this

8.

ALGEBRA.
"Firft, take care
to

45
difpofe

manner.

the

terms according to the dimenfions of the alphabet^

as in divifian

then find the fquare root of the fir ft

firft

term aa, which gives a for the

root.

member of the

Then fubiratt

its

fquare from the propofed

firft

quantity',

and divide the

term of the remainder

(lab

-\-

z b ) by the double of that member^ viz. 20,

and
root.

the

quotient

is

the fecond

member of the
the double of
-j- b) by the

Add

this fecond

member

to

the

firft,

and multiply

their

fum (2a

fecond member b, and fubtracl the producl (lab

z z -f b ) from the forefaid remainder (lab -\- b ) and if nothing remains, then the fquare root is

obtained

and

in this

example

it

is

found to be

b.

The manner

of the operation

is

thus,

z a *Jr 2ab-{-b {a-\-b

ia-\-

bxiab-\-b z
b'lab -\-b z
o

But if there had been a remainder\ you muft have divided it by the double of the fum of the

two parts already found, and the quotient would have given the third member of the root.

Thus if the quantity propofed had been z a*-\- 2ab+2ac + b + ibc -f- c\ after proceeding
as

above you would have found the remainder

zac

46

Treatise

of

Parti.

2ac+2bc + c\ which divided by la + ib gives c to be annexed to a^- b as the 3d member of the Then adding to 2a-\--ib and multiplyroot. ing their fum 20 -j- ib-\-c by r, fubtraft the pro r
% duct 2ac-\-2bcAr c from the forefaid remainder

and

fince

nothing
c is

now

remains, you conclude

that a

+b+

the fquare root required.

The
z

operation
z

is

thus
z

a -{-2ab+2ac

+ b + 2bc-\-e
z

{a-\-b-\-c

2a

+ b\ 2ab + 2ac + b + 2bc -f

Xb
2ab
-\-

+b
c

la

+ lb

\2ac -\- 2bc-\-c

X f/2^f+ 2^+t*

Another Example.
##

<z# r-j-

aa

(x

--ia

AW
ix

X ax
-J**'

ia\ ax

-J-

-\-

-J^ i#a

The
100,

fquare root of any

after the
its

fame manner.

If

number is found out it is a number under

is

neareft fquare root

-

found by the
its

folis

lowing Table

by which

alfo

cube root

found

{Z/iften .fta,

4.

&

TAB."Vm

>W

4-f

Chap.
found
it

8.
if it

ALGEBRA.
be under 1000, and
its
1
I

47
biquadrate if

be under 10000.
2
3

The Root
Square

4
16

25

Cube
Biquad.

8 27

64 125

36 216

49
343

64
512

9 81 729

|i

16 81 256 625 1296 2401 4096 6561

is

But

if

it

fquare root will

number above 100, then its confifl of two or more figures,

different operations

which muft be found by

the following

by

RULE.
in the -place of units, pafs

53. Place a point above the number that is the place of tens, and

place again a point over that of hundreds,

and go

on towards the

left

hand, placing a point over every

2d

figure

and

by thefe points the

number will be

difiinguifhed into as
in the root.

many parts as there are figures

root of the firfi

Then find the fquare

part, and

it

will give the firfi figure of the root

fubtracl

its

fquare from that part, and annex the

fccond part of the given number to the remainder.

'Then divide this

new number

(neglecling its lajl fi-

gure) by the double of the

firfi

figure of the root,

annex the quotient

to that double,

and

multiply the

number

thence arifing by the faid quotient,

is lefs

and if

the produil
it,

than your dividend, or equal to

that quotient fhall be the fecond figure of the

root.

But

if the product
lefs

is

greater than the di-

vidend, you muft take a

number for the fecond

figure

48
the fame

^Treatise?/'

Part

figure of the root than that quotient.

Much

after

manner may the other figures of the quotient be found, if there are more points than
two placed over the given number.

To
point

find the fquare root of

it

9985 6 i

firft

99856, then I find the fquare root of 9 to be 3, which therefore is the firft figure of the root; I fubtract 9, the fquare of. 3,
thus,

from

9,

and

to the remainder I
I

annex the fecond

part 98, and

divide (neglecting the lafl figure

I

8) by the double of 3, or 6, and

place the
1$

quotient after

-6,

and then multiply 61 by

and fubtract the product 61 from 98. Then to the remainder (37) I annex the lafl: part of the propofed number (56) and dividing 3756 (neglecting the
that
is

laft

figure 6)

by 62,

by the double of 31* place the quotient after, and

I

multiplying 626 by the quotient 6,

find the

product to be 3756, which fubtracted from the dividend and leaving no remainder, the exact
root rnuft be 316.

EXAMPLES.
99856(316

9_
61V98 xiJ6y

626^3756
*6< 375 6

Chap.

8,

ALGEBRA,
5*9
(

49
23

2 739475 6 (5234
.

25

4
43\* 2 9 X3/129

J02A229 X2> 204

1

43 j3547
>

*3

3 ]2 9

10464 41856

*4'4 l8 5 6

54 In general, to extract any root out of any given quantity, " Firft range that quantity
according to the dimenfiom of
its

letters^

and exand that

required.

tract the faid root out of the firft term,

ft? all

be

the firft.

member of the
that denominates

root

*Ihen raife this root to a dimenfion lower by unit

than the number

quired^

the

root

re-

and

multiply the
-,

power that

arifes by that

number itfdf

divide the fecond term of the given

quantity by the pre duff,

and the

quotient fh all give

the fecond

member of

the root required"

Thus

to extract the root of the 5th

cf *4- $afy

\oa>b

ioa

fc

power out b\ I $ab*

find that the root of the 5th

gives a, which I raife to

power out of a s the 4th power, and

is

multiplying by 5, the product

$a A

then di-

viding the fecond term of the given quantity a 4b

by 5# 4 ,

find

b to be the fecond

member
and

jo
and
ing

^Treatise of Pa*t I. railing a + b to the 5th power and fubtractit,

there being

no remainder,

conclude that

b is the root required. If the root has three a members, the third is found after the fame manner from the firft two confidered as one member, as the fecond member was found from
the
firft ;

which may be
faid

eafily

underftood from

what was

of extracting the fquare root.

5$. In extracting roots it will often happen that the exact root cannot be found in finite
terms
to be
\

thus the fquare root of a x

-J-

xz

is

found

^2a

Sa^i6as
The
1

128*7+'
is

<SC

operation
v

thus
X*
s
l

~ 2*

X6 i6# 5
'

8a

2^4-

)-*4-ff-

**
1

*4

<*

8*V~"

4a a

8**

*""

64a 6

8**

64a6

After

Chap.

8.

ALGEBRA*
the

ft
cube root of

After the fame manner,

a1

-f-

tf will be found to be

8 243a" 3^ 9 T 8ia 56. " The general Theorem which we gave

'

'

C9V "

for the Involution of binomials

will

ferve alfo

for their

Evolution

becaufe to
is

extract

any

root of a given quantity

to
raife

the fame thing as

power whofe exponent is a fraction that has its denominator equal to the number that exprcffes what kind of rooE
that quantity to a
is

to be extracted.

Thus*
to
:

to extract the fquare

foot of a

+ h +mX
&c.

is

raife

~f-

b to a power

whofe exponent

is 4-

Now

fince a-\- b

= a m

mX

~ am b
l

" a m i b t 42

mx

Xa m ~&

23
X
b
b
%

Suppofing

==

4a

y ou will ^ n ^

XiXia
b*
:

-J.

JL

*b\ &c.

== **

+
2**

7
$a z

Csf^r.

And

after this

manner you

will

find that

11
fore*

2#

btf

'

ioa s

57.

The

roots of

numbers

are to

be ex-

tracted as thofe of algebraic quantities.

" Place
every

a point over the

units,

and then place points over

52

A Treatise
hand> according as
it

of

Part

I.

every third, fourth or fifth figure

left
is

towards the

the root of the cube,

is

of the 4-th or th
there
he

power

that

required
to

*,

and

if

any

decimals

annexed

the

number

point them after the fame manner', proceeding from

the place of units towards the right hand.

means the number will be divided

riods

into fo

By many

this

pe-

as

there

are figures
is

in

the

root required. cube,

'Then

enquire

which

the

greateft

biqua-

dratty or $th

power

in thj firfi period,

firfi

and the root

of that power will give the

required.

figure cf the root

Subtract the greateft cube, biquadrate,

fir period, ft

or th power from the

and

to the re-

mainder annex the

firft figure

of your fecond period^

which Jhall give your dividend.

Raife the
lefs

firft

figure already found to a

power

by unit than the

is,

power whofe

root

is

fought,

that
as
it

to

the

2d, 3 d, or 4-th power, according

is

the cube root, the root of the ^th, or the

root

of the

th power thai

is

required,

and mul-

tiply that

power

by the index of the cube, ^th or

this producl,

^th power, and divide the dividend by

fo Jhall the quotient
root required,

be the fecond figure of the

Raife the part already found of the root, to

the

power whofe

root

is

required,

and

if

that

power be found
right.

lefs

than the two

firft

periods cf
is

the given number, the fecond figure of the root

But if

it

be

found greater, you muft

till

di-

miniflj tfa

fecond figure of the root

that power
be

Chap.
be

8.

ALGEBRA.
to or kfs

53

found equal

than thofe periods of the

it,

given number.

Subtract
5

and

to

the remainder
till

annex the next period

and

'proceed

you have

gone through the whole given number, finding the

^d

figure by means of the two. firft, as you found

,

the fecond by the firft

and afterwards finding the

4th figure (if there be a \th period) after the

fame manner from

the three firft"

Thus
it

to find the cube root of

138x4

point

13824*, find the greateft cube in 13, viz. 8,

is

whofe cube root 2

required.

the

firft

figure of the root

Subtract 8 from 13, and to the re8 the

firft

mainder 5 annex
period
viz.
\

figure of the fecond

divide

58 by

triple
is

the fquare

of 2,
the fe-

12. and the quotient

4,

which

is

cond

figure of the root required, fince the

cube

of 24 gives 13824, the number propofed. After the fame manner the cube root of 133 12053 is found to be 237.

OPERATION.
13824 (24
Subtr.
S

3x4 =

12) 58 (4

Subtract

24x24x24=13824
Rem.
o

13

^4

Treatise
i33i 20 53
8
(

of

Part

I.

= 2X2X2
=

2 37

f?) 53 (4 or) 3 Subtract 12167 23 x 23

23

3x23x23=1587)

1450 (7 Subtract 133 12053 == 2 37

1

237

x 237

Remain.

In extracting of roots, after you have gone

through the number

remainder, you

propofed,

if

there

is

may

continue the operation by

adding periods of cyphers to that remainder,

and

find the true root in decimals to

any degree

of exaclnefs.

CHAP.
58.

IX.

Of Proportion.

II7HEN quantities of the fame kind VV are compared, may be

it

conii-

dered either

how much
is

the one

is

greater than
or,
it
is

the other, and what

their difference-,

may

be confidered
the

how many
other
5

times the one

contained in

or,

more

generally,

what

Chap.
what
tio

9.
is

ALGEBRA.
their
is

$5
relation

quotient.

The

firft

of

quantities
\

expreiled by their Arithmetical ra-

the fecond by their Geometrical ratio.

ratio
is

That
is

term whofe
antecedent^

enquired into
it is

is

called the

and that with which

compared

called the ccnfquent.

59.

When

of four quantities the difference

betwixt the

firft

and fecond

is

equal to the dif-

ference betwixt the third and fourth, thofe quantities

are called Arithmetical proportionals

3,
e,

as the

numbers
a,

7,

12,

16.

And

the quantities,

a-{-b,

e-\-b.

Rut
fame

quantities

form

a feries
increafe

in arithmetical proportion, or dccreafe by the

thefe,
a,
b,

when they "

-\-

confiant difference."

As
&c.

a x

x9 x

+ b,

-j-

2#, a

3b, a -f 4^,

ib*
io,

&c. or the numbers,

7,

1, 2, 3,

4, 5,

&c. and

4,

1,

2, 5, 8,

60. In four quantities arithmetically proportional,

<c

the

fum

of the extremes

is

ecual to the
<?,

fum
are

of the

mean terms"

Thus
is

a, a-\-b y

+ b,

arithmetical proportionals, and the

the extremes (a

+ e + b)
(a-\-b-{-

equal to the

fum of fum of

the

mean terms

e).

Hence,

to find the

fourth quantity arithmetically

proportional to

any three given quantities and

third,

and from

their

" Add the fecond fum fubtract the firft:

term, the remainder fhall give the fourth arithmetical proportional required."

E3

61.

56
6:. In a
<c

A
/i?

Treatise

of

Part

fcries of

arithmetical proportionals

the

fum of the fir -and laft terms is equal to ft fum of any two terms equally diftant from the
If

extremes"
a-\- ib>

the

firft

terms

are

a,

-\~

~b9

&c. and the

but three

laft

term
laft

x, the laft

but one will he

the
laft

xb,
x

the

but two x

term
2b,

jj,

&c.

So that the

firft

half of the terms, having thofe that are equally

diftant

from the
;

laft

term

fet

under them, 'will

frand thus
a,

x9

a-\r b,

4-2* *M~3* *

b9 x

+ 4*> 2^, x 3#, x 4b, &

~\-

a-\-% 9 a-\-x 9 a

x,

-j-

x,

-\- x,'

&c.

And

it

is

plain that if each term be added to

it,

the term above

to the
x.

the
firft
is

fum

will

be a

+ x equal
laft

fum of

the

term a and the

term
equal

From

which.it

plain, that

"

the

fum of
is

all the terms of

to the

an arithmetical progrefjion

fum of

the firft

and

laft

taken half as often

as there are terms" that

is,

the

metical progreflion
firft

is

equal to the

and

laft

terms multiplied

fum of an arithfum of the by half the num-

ber of terms.

n be the

Thus in the preceding feries, if number of terms, the fum of all the

terms will be a
62.

+ x X -. 2
difference of the terms
in

The common
is

being
term,

3,
it

and b not being found

plain that

the
in

firft

"

its

coefficient

any
term,

Tab tx

^u? 44

Chap.
term

9.

ALGEBRA.
Therefore
in the lafb

$y
term

will

be equal to the number of terms that

precede that term.""

x you mufl have

equal to

a + n

the terms being

n\ x b, fo that x mufc be X b. And the flim of all a-\-x X ~ it wHl alfo be equal
1
5
,

7.an-\-nH
to

to

nb
,

nb

cr to a A
1

X #

Inusror

example, the
continued
2

feries

X 100

4-

0000

ico =
fum

hundred,

5> &t* 4 3 mull be equal to

5050.
its

63. If a feries have (o) nothing for

fird

term, then

st

its

fimll be equal to half the

product of the

Iaft

term multiplied by the number

then,
is

of terms"

For

a being
in

o, the

fum

of the terms,
will in this cafe

which
be

general

a-\-xx,
it

From which

is

evi*

den^, that " the

fum of any number of fum of

as

arith-

metical proportionals beginning from nothing,

is

equal to half the

many terms

equal

to the greated term.

Thus + + 2 +3 + 4 + 5 + 6 + 7 + 8 + 9=: 9+9+9+9+ 9 +9+9+9+ 9+9 __ I0X 9 _ r

1

~ 45
,

64.

"

If

of

fo'jr quantities

the quotient of

the

firft

and fecond be equal to the quotient of

the third and fourth^ then thofe quantities are

laid to

be in Geometrical proportion"

Such arc
the

5
the
&,

A T R E AT
2,
-,

I S

E of

Part

I,

numbers
ar,
b,
-,

6, 4,

12

and the quantities

after

br

which are exprefTed

this

manner

2:6:
a
:

4
b

12.
br.
is

ar

::

And you -read them

}s

by faying, As 2
is

to 6, fo

to 1 2

or as a

to ar, fo

is

b to br.

In four quantities geometrically proportional,

fl

the product of the extremes

is

equal to the product

of the middle terms?*

Thus a x

br

= ar X

And,

if it is

required to find a fourth proportion

nal to any three given quantities,

"

multiply the

fecond by the third, and divide their produff by the

firft,

the

quotient JJoall give

the fourth propor^

tional required"

Thus,

to find a fourth
I

pro-

portional to a, ar, and b,

multiply ar by b,
firft

and divide the product ark by the

a, the quotient br
is

term

the fourth proportional re-

quired.

6$.
little

In calculations

it

fometimes requires a
;

care to place the terms in due order

for

which you may obferve the following Rule. " Firft fet down the quantity that is of
fame
ftder,

the

kind with

the quantity fought,

queftion,
lefs

then con-

from
is

the nature of the

is

whether

that which

given
\

is

greater or
is

than that

which
greatefl

fought

if

it

greater, then place the

of the other two quantities

but if
it is lefs,

on the

left

hand
two

-,

place the

leaft of the other

quantities on the left

hand, and the other on

th$

Chap,

9,

ALGEBRA.
Then
fhall the

59

the right."

terms be in due or-

der

-,

and you

are to proceed according to the

rule,

multiplying the fecond by the third, and

firft.

dividing their product by the

EXAMPLE.
If 30 men do any piece of work in iz days,

how many men

Becaufe
it is

fhall do it in

8 days ?

number of men that is fought, number of men that is given I eafily fee that the number that is given than the number that is fought, is greater therefore I place 18 on the left hand, and 12
a
I firft fet
:

down

30, the

on the
o l8, 30,

right

and find a fourth proportional to

J2, VIZ.

3 ^j

* 12

20.

66.

When

a feries of quantities increafe

by

one

common multiplicator, or common divifor, they are faid

As
a, ar, ar
1 ,

decreafe by one
to be in

" Geo-

metrical proportion continued."

ar 3 9 ar+, ar 5 , &c.
a 7T' a 7*' a 7T'

or,

a
?'

>? The common

their

a 7^

&$'
is

st

multiplier or divifor

called

" common

ratio."

In fuch a
loft
is.

feries,

"
or

the product of the firft

to

and

always equal
but one,

the produft of the fecond

and

loft

to

the produtl
the
if

of any two In the

term,

terms equally remote

feries

from

extremes.

a t ar, ar z P ar\ &c.

be the

laft

then

6o

i
'jri

Treatise
nowit
r

Parti,
feries

then mall the four lad terms of the

y,

be

-&* jt*

is

F laln that
i

ax y -

% 6j. " I'he fum of a feries of geometrical proft term, is equal to the portion ah wanting the fir

fum of all but mon ratio"

the loft term multiplied by the com-

& + 4 + -+X y y y_ = rx^ + ar + r% &c. 4- + i + +

For arffrXt* + Ar', &c.
-ji

Therefore
will
sr

if

be the fum of the

feries,
is,

be equal to
or sr

syycT'\
-=yr

that

a
-=.

jr,

a,

and s^-r-^-p
is

68. Since the exponent of r

always in-

creafmg from the fecond term,

if

the

number

of terms be

, in the laft

term

its

exconent will
\

be n

i.

Therefore y :b ar n
,

~~ x

and yr
ar
r
n

=
So

^ r x + i

^;-

ana
firft

/)'?*

Vr

)=
a\
1/
feries,

a
.

that having the

term of the

the

numyou
to be

ber of the terms, and the

common
all

ratio,

may
If

eafily find the

it
is

fum of

the terms.
is

a decreafing feries
-f-

whofe fum
-

found, as of y
ar*

-f % -f 4' ^ c

~f- &?*

-f

+ ar

-j-

a^

and the number of the terms be

* See the Rules in the following Chapter.

fupr

Tab x

irk

^w

+#

Chap. io.
fuppofed

ALGEBRA.
For, becaufe
n,
is

61

infinite,

then mall a^ the lad term, be

equal to nothing.

and confe-

quently r x

infinite,

= -^ = o.
r
n
~

The fum

of fuch a

feries s

= r

which

is

a finite fum, '

though the number of terms be

infinite.

Thus
and

1+i+ L+ i+ ^+,
i

&c.

=~ =

2.

-K+i.+^+^+, f.

=j^ =

4.

CHA

P.

X.
only

Of E q^u ations that involve one unknown Quantity.

6g. J\

equation

is

" a
of

proportion offer iing

quantities."
It

JljL
is

the equality

tkxio

expreffed moil commonly by

fetting

down

the

quantities,

and placing the fign (=) between

them.

An
when

equation gives the value of a quantity,

that quantity
:

is

alone on one fide of the

is

equation
that are

and that value

known,

if all thofe

on the other
x

fide are

known.
a

Thus

if

I find that

=
3

=8,1 have
laft

known
we

vaare

lue of *.

Thefe are the

conclufions

to

62

Treatise

of
;

Parti.
and
if

to feek in queftions to be refolved

there

be only one unknown quantity

tion,

in a
it,

given equafuch a value

and only one dimenfion of

may

always be found by the following Rules.

RULE
70.

I.

" Any
9i

quantity

may

be tranfpofed from one

fide of the equation to the other, if you change

its

fgn.
to take
it

For

away a quantity from one from both

fides

fide,

and to place

with a contrary fign on the other

,

fide, is to fubtract it
is

and

it

certain, that

" when from

equal quantities you

fubtract the

fame quantity, the remainders

mud

be equal."

known and unknown may quantities are mixed feparate them by bringing all the unknown to one fide, and the known to the other fide of the

By

this

Rule, when the

in

an equation, you

equation

as in the following

Examples.

Suppofe

5v

-J-

50

= 4x -\-561
or,

By

tranfpofit.

And

if

4^=56 50, ix a = x 2x x = b RULE

5*
-j-J-

x^-6.

b9

#, or,

x =zb

a,

II.

71. " Any

tity
is

quantity by

which the unknown quanbe taken

multiplied

may

away, if you,

divide all the other quantities on both fides of

the equation by //."

For

Chap.
For
tion

io..

ALGEBRA.
is
,

63

that

to divide both fides of the equa-

by the fame quantity and when you divide equal quantities by the fame quantity, the quotients

muft be

equal.

Thus,

If ax

b,

then

#=:

b
a

And
by Rule
and by Rule

if
1.

3#+
x ==

3^ = 27 12=115,

12

= 27,

2.

== 5.

Alfo

if
1.

ax

2&*

by Rule
and by Rule

ax=z%cc
x

iba
2&

srr,
9

2.

"ice

RULE
72. If
the
quantity, that quantity

III.
is

unknown quantity
all the other

divided by any
if

may be taken away

you multiply
tion by it."

members of the equa-

Thus,

then fhall x

= bb + $b.

If

4-- 41=10,
5

#
and by Rule
1.

+ 20 = 50, #= 50-20:= 30.

If

64

A
If

Treatise

of

Part

I.

72

-f-

4ff'-f

by Rule and by Rule

1.

18 = 6x ^-^Xyox^
#

= 2X + 6, 72 = 6# +18,
24

ix.

2.

= = 27.
is

By
mall

this

Rule an equation, whereof any part

a fraction, be

may

be reduced to an equation that

expreffed

by

integers.

If there

are

more fractions than one in the given equation, you may, by reducing them to a common denominator, and then multiplying
all

the other

terms by that denominator, abridge the calculation thus;

If

*
5
3

*~^ n
3 *

*"

then

+5

-= x
1

7,
1

and by this Rule 3 x

+ $x =

5#

05,

and by Rule

and

2.

^=^-5 = 15.
IV.

RULE
73. "If
volves the

that member of the equation that in-

unknown
is

quantity be

a furd

root,

then the equation

to be reduced to another

that fhall be free from any furd, by bringing

that member

firft

to ft and alone

upon one fide

of the equation,
radical jjgn

and then taking away the

it,

from

and rcifmg the other

fide of the equation to the

power denominated

by the furd:\

Thus

IAS,. XT.

&&

S3

^V- S4

ty

Jo

Chap.

10.

ALGEBRA.
if

65

Thus

v^4^+

16

z=.

12,

and
,

4#+ 16= 144, 4#= 144 -*- 16=

x

and

=. 128 z

If v/^a; -\- b

= 4
c
:

28*

32. D

d>
9

then

/ *#~

-f b

=d + c

ax

-\-

bz

=zd 2 + idc

+ <;%
~.

and

# ==

If

z \/ a x

b x=a
z <z
?

then

a x * 2 # ==
l

and

=
V.

RULE
74.

" If

that fide of the equation that con-

tains

the

unknown quantity be a
or other

compleat

fquare,

cube,

power
root,

then extracl
the
root

the fquare

root,

cube

or

of

that power, from both fides of the equation*

and

thus the equation fhall be reduced to one of'

a lower degree"
If x
z

+
-J-

6x
3

~f-

= 20,
\/ 20,

then x

=+

and x =s

v/20

3.

If

66

A
If x *

TRtATlst

df

PartL

+ ax + fL - *, 4

then

a? -J

=+

b<

and #
1

+ =

-.

If x

-]-

14*4-49
*
1 1

= 121,
5

then x

and

a?

+ 7=+ = + 7 = 4, or 18*

RULE
" 75.

VI.
be converted into an

proportion

may

equation, ajferting the producl of the extreme

terms equal to the producl of the mean terms %

or any one of the extremes equal

duel of the means divided by

to

the pro-

the other ex*

treme"

x 4 then 12 #= 2#....3# = Or 20 x x 7 then 60 3# = jx icw = 60

If 12
:

i, *

12.... and

#=4.

if

3,

....

...

and #

= 6,

RULE
76.

VIL
found on both
fides of

" ^f

^jry quantity be

the equation with the

fame

fign prefixt, it

may
or

be taken
quantities

away from
in

both :"

"
are

Alfo 9 if all the

multiplied

the

equation

divided

/i&n.fut

TAB.XK

&

S<P

Chap. io.

ALGEBRA;
it

67
may
be ftruck

divided by the fame quantity*

out of them all"

Thus,
,

If %x

+ b~a + ...3*=#....and x ==
8r..and^=-

If sax+5abz=. Sac,. ^x~\-^b=.

=~

RULE
77.

VIII.
in

" Injkad

of any quantity

an equation

you may

fubftitute another

equal to it?\

Thus,

if

3*+j = 24,

and
then

=9 3* + 9 = 24
y
;

. .

# :=

4 ""~9 _. g,

If

3^+5^=
j=5^j

I2 >

and
then

15* -j- 5^ (=s 20#)

#

and

=
20

= 120,

6.

The

further
in

improvements of

this

Rule mall

be taught

the following chapter.

M A P.

68

*4

Treatise

of

Parti.

CHAP. XL
Of
the Solution

of queftions that

produce fimple equations.

SIMPLE equations are thofe "wherein the unknown

quantity
is

only of one dimenfion

:'*

In the folution of which

following directions.

we

are to obferve the

DIRECTION
78.

I.

After forming a

diflincl idea

of the que-

Jiion propofed, the

unknown
and

quantities are to be

expreffed by letters^

the particulars to be
into the
is y

tranjlated

from

the

common language

algebraic
into

manner of expreffng them 9 that

fuch equations as fhall exprefs the relagiven of fuch quan-

tions or properties that are

Thus,

if the

fum of two
is

quantities

that condition
If their

expreffed thus,

mud be 6o, x+y = 6o.

gives
If their

muft be 24, that condition x y==24>. product muft be 164.0, then xy 1640.
difference

If their quotient
It their

muft be
is

6,

then

6.
: :

proportion

or 2x

2y

as 3 to 2, then

2,

becaufe the product of the

ex-

tremes

Chap.

ii.
is

ALGEBRA.
equal to the product of the
II.

69
mean

tremes
terms.

DIRECTION
79.

"

After an equation

is

formed, if you
then, by the
it to ft and

have one unknown quantity

only,

Rules of the preceding Chapter9 bring

alone on one fide, fo

as to have

only

known

quantities on the other fide :"

thus you fhall

difcover its value,

EXAMPLE.
A
was
It

perfon being

afked

what was

his age, an*

fwered that ~ of
his age?

his age multiplied

by

of his

age gives a producl equal to his age,

Qu. What

appears from the queftion, that

IX

if

you

call

his

age x, then &

(hall

X == #> 12 4
-~- =3 x
40
;

that

is

and by Rule and by Rule

3.
7.

%x z

whence by Rule 2.

... ...

3* x

= = =

48.V,

48,
16.

DIRECTION
conditions of the queftion
tities
:

III.

^ 80, " If there are two unknown quantities, then there muft be two equations arifing from theSuppcfe the quan-

x and y

find a value of x or y,

from
each

yo
each
thefe

A
of
will
arif'e

Treatise
to

of

Part

I.

the equations,

and then by putting

each
other,

two values equal

quantity

there

a new equation involving one un\

known

which muft be reduced by the

Rules of the former Chapter*

EXAMPLE

I.

Let the fum of two quantities be s, and Let s and d be given, and let difference d.
elves. required to find the quantities themf

their
it

be

Sup-

pofe them to be x and y, then,

pofition,

by the fup-

x y = d
~~
"~""^
7

jy == s

whence <
and

Lx d -\-y d + y = y iy = d
s s

"

s
A and x
S

=.

d +

EXAMPLE
Let
it
s,

IL

be required to find

two numbers whofe

as a to
b.

fum

is

and

their proportion

Let

the numbers be

x and y, then fhall

Suppof.

Chap. ri.

ALGEBRA,
Sx

71

Suppof.

+? = s
:
:

t x :y

bx r= ay
ay

X .=

?
J

?=-7
?*=
ay
-\-

& a + b xy = &
by

==

bs

y "~ a

b
<75

'

-f

EXAMPLE
/f privateer
<?#

III.
;//<

running at the rate of 10

2wr,

covers dif

a fhip 18 mk&j
tf#r

0^"
:

making
de-

way

at the rate of 8 z*7 #

miles

It is

manded how many

fhe be overtaken ?

the fhip can

run before

before ?at be overtaken be called x\

Let the number of miles the fhip can run and the

fhe

number of miles the privateer muft run before come up with the fhip, be y then mall (by Supp.) and x y 8 io, y =# + 8
.

A-y

-,

whence

ox

= 8j

5 3

. ,

and x y

Whence

./4

Treatise

<?/"

Parti.

Whence y

18=-^,

and ^=-90 .Xx'zzy

18
;

= 72.
To
find the time, fay, if 8 miles give
1
: :

hour

.72 miles will give 9 hours.

Thus> EXAMPLE IV.

8

72-: 9.

Suppofe the diftance between London and Edinburgh to be 3 60 miles, and that a courier fets out from Edinburgh running at the rate of 10 miles

an hour

another fets out at the fame time from

London, and runs 8 miles an hour. It is required to know where they will meet ? Suppofe the courier that fets out from Edinburgh runs x miles,
and the other y miles before they meet
by fuppof.
x +jy =f 36 x :y:: 5 4
:
-,

then fhall

360

4
9^ =3 1440 '44 ^y

60

== 200. EXAMPLE V.
# z= 360
j.

Two perfons A, if B sew^/i

difcourfing of their revenueSy fays

jy/WJ

him a

poft he has of

25

1.

year,

Chap., ii.

ALGEBRA.
:

73
Says B, if

year, their revenues would, be equal

A would give

him a place he

holds of

22

1.

per

annum, the revenue of B would A. Qu. their revenues ?

Let the revenue of

be double that of

A be called
5

x> that

of B,

-,

then,

by fupp.

2"'~^5-~y~
y =?
j
2.v

66 =
=

*
2.V

+ 25 44 +
x

25
2
2

=
=r

# -f 50
2,V

66

# r= 66 J- 50 =116 a -L 50 166. J

EXAMPLE
people,
to each

VI.

^ gentleman diftributing money among fome poor

found he wanted 10
;

s.

to be able to

give 5 s.

therefore he gives each

s. left,

s.

only,

and finds

that he has 5

Qu.

the number of /hillings

and poor people ?

Call the

number of
,

the poor x, and the

num-

ber of millings y

then,

by fupp.

S& jy

+ *o 5 = 5* 10
y
5

> = 4* + 5 5* 10 =-4* + 5* 4*= 15

#
;y

= =

15

4*

+5=
E
4

%.

E X-

74
Two
flock

Treatise

of
VII.
\

Parti.

EXAMPLE
merchants were copartners

thefum of their

was 300 1. One of their flocks

1 1
-,

continued in com-

pany

months* but the other drew out his flock in

9 months
flock?

when

they

made up

their accounts they

divided the gain equally.

Qu. What was

firfl
\

each man's
to be >\

Suppofe the flock of the

and the flock of the other to be y

then,

=. 300 y ij4-5(7 = 33 00
x
-=,
,

2oj

= 3300

EXAMPLE
x and y
fupp.
-

VIII.
is

2*&r* are two numbers whofe fum

the 6th part

of their produff 9 and the greater is to the lejfer as 3 to 2. Qu. What are thefe numbers ? Call them
9

then,

*J
:

'

~6
:

I x

7
#

7-"

= 3^ 3qy = 3iv
12^
r

6"

1 S_y

* 6xx = oy
6v

=- =
3

io

3X10
2

^6

= V* whence*

PI

Chap.

ii.

ALGEBRA,
DIRECTION
IV.

75

81.

" When

in one of the given equations* the

is

unknown

quantity

of one dimenfion* and in

*

the other of a higher dimenjion

you muft find

a value of the unknown quantity from that equation where it is of one dimenjion* and then
.

raife that value to the -power

of the unknown
*

quantity in the other equation

and by com-

paring

it*

fo involved* with the value you de-

duce from that other equation* you Jhall obtain

an equation that will have

quantity*

only one

unknown

and

its

powers"

That

is,

when you have two equations of difyou muft

ferent dimenfions, if you cannot reduce the higher

to the fame dimenfion with the lower,

raife the

lower to the fame dimenfion with the

higher.

EXAMPLE
The
[urn
their fquares* being given*

IX.

of two quantities* end the difference of

to find the quantities.

Suppofe them to be x andjy,

difference of their fquares d.

their

fum Then,

s*

and

5*
'

lx y
z

4-jy
z

=
z

~d
%

*2syz=zs z
'*

x
x
l

^sy 2sy+y
s

d
d

+ y ~ 2sy~Ly* = 2sy* whence

z

r=d^y\
s
z

* -~zr *+ d and* = -_.

y
s

EXAMPLE

76

Treatise

of

Parti.

EXAMPLE
Let the proportion of
'tzvo

X.

numbers and the fum

let it

of their fquares be given,

and

be required to

find the numbers th em elves.

Suppcfe their pro-

portion to be the fame as that of a to b y and

let

the

fum

of their fquares be

-,

that

is,

let

x
-{
%

b
c

x 4- y

=:.

then x

= ~> b
-~
b
%
y

and x z z=.
but #
4

= c j%

whence

y
y
z

-=.

-~z
:

ax

+ a y =z cb + b xy* = cb
z

y ~~

+P

J^'dx =J^<
XL
their
difference of

EXAMPLE
a
to

Let the proportion of two numbers be that of

by

and the

cubes be d*

Qu. What

are the numbers?

Then,

j?

Chap.

n.

ALGEBRA.
{x :y x
X
ay =.-,
,
:
:

77

x
but

and *
?

a = -jj-

s s -3^3 y

a^=

whence
and a^y
1

J+j = ~^-, i +
3

jy

fry z = db %

aV ^__ ~~ a 3 P

y- J-J!L.
3

and*

= /

daZ
.

DIRECTION
82. u If
there

V.
quantities*

are

three

unknown

there muft be three equations in order to deter-

mine them, by comparing which you may, in

all cafes,

find two equations involving

quantities
thefe
\

only

two unknown
tion
3,

and

then, by Direc-

two you may deduce an equation involving only one unknown quantity ;

from

which may be refolved by the Rules of the Chapter^

loft

From
unknown

three

equations

involving any three

quantities, x, y,

and z, to deduce two

equa*


A
Treatise
Rule

y
tities,

of

Parti.

equations involving only two

the following

unknown quan-

will always ferve.

RULE.
" Find three values of x from
tions
;

the three given equa-

then, by comparing the firfi

and fecond

value-,

you will find an equation involving only

-,

y and z
thirds
only

again* by comparing the

-"
.

firfi

and

you will find another equation involving

y and z

and

laftly*

thofe equations are

to be refolved by

Direction 3.

EXAMPLE
Suppofe

XII.

&

~ y~~ z l + y+ %=i2i f i2 (then, J 20 ly 3% +2? +32; = 20

(

j ft

2cQ 3

12

2;= 32
20-

2j

Thefe two lad equations involve only j and 2, and are to be refolved, by Direclion 3, as
follows.

%y

Chap.

ii.

2jy

ALGEBRA. 7+322 = 26 12=8

3jy

i J+22=8

62= 24 2y 22 12= 7+42 whence y = < 42 2d 22 = 12 42 22= 12 8=4

^6
8
2
22;
. .

ft

value. value.

and

2=2
jy(=

8 22) = 4

#(=i2
83. This method
tend to
all
:

y =
2)
eafier

6.

is

general, and will ex-

equations that involve three

unknown

quantities

but there are often

and fhorter

unknown

methods to deduce an equation involving one quantity only ; which will be bcft

learned by practice.

EXAMPLE
#4- y r*4-J
Suppofing ix \ x

XIII.

y = 4
6

+2=

26

by addition 3*
1

= 36
36 = 12
3

x=.

y =z X

4=8
6=6.

EX-

8o

4
x

Treatise

of

Parti.

EXAMPLE
Suppofing
I

XIV,

+ya x +z=h 1^ + * =
y a y -\-zz=.b
x-=.a

+ zzn c a -j-o +2z b + c

y
iz-=-b-\-c
b

+ = ca 2

<'

84. It
rections, in

is

obvious from the 3d and 5th Diif

what manner you are to work

there

are

four, or more,

unknown

quantities,

and

four, or

more, equations given.

By comalways at

paring the given equations, you

may
is

length difcover an equation involving only one

unknown
equation,

quantity-,

which,

if it

a fimple

may

always be refolved by the Rules

of the
that

laft

Chapter.

We

" "When

there are as

may conclude then, many fimple equations

thefe

given as quantities required,

quantities

may

be difcovered by the application of the pre-

ceding Rules.''

85.

Chap.

12.

ALGE

A.
more

81
quantities

85. "
ftion
is

If indeed there are

required than equations given,

then the que-

not limited to determinate quantities

capable of an infinite

but

is

number of
it

folu-?

tions."

And, " If there are more equations

given than there are quantities required,

may
be-

be

impoiTible to find the quantities that will

anfwer the conditions of the queftion 5"

caufe
fiftent

fome of

thefe conditions

may be

incon-

with others.

CHAR
for

XII.

Containing fome General Theorems

the exterminating

unknown

Quantities in given Equations.

we call thofe co" fame order" that are preefficients of the fixt to the fame unknown quantities in the dif-

IN

the following Theorems,

ferent equations.

Thus,

in Theor. 2.

a> d,

gy

are of the fame order, being the coefficients of

alio , e, , are

of the fame order, being

:

the coefficients of

and thofe are of the fame

order that affect no

But thofe are

unknown quantity. called " oppofite" coefficients

from a
different equation,

that are taken each

and


A
J,
e9

82
and from a a 9 e 9 and
?,

Treatise
of
firft

of

Part

different order

coefficients

As,
,

b 9 in the
>

Theorem
a>

and

ky
ky

in the fecond
Sec,

alfo,

h9

and

d>

by

THEOREM
involving two

I.

86. Suppofe that two equations are given,

unknown
ax

quantities, as
=.

+ by
-L-

{ dx

ry

f
f do
L.

then mail y :=

^~"
ae

Where

the numerator

is

the difference of the

in

products of the oppofite coefficients

ders in which

the or-

is

not found, and the deno-

minator

is

the difference of the products of the

oppofite coefficients taken from the orders that

involve the two

unknown
firft

quantities.
it is

For, from the

equation,
.

plain that
-,

ax

= c by
/ ey
d
,

and x
and x

=
a

from the 2d, dx ==

therefore
c ~~

= Jf

gy
,

=.

and*: J

whence

aey

dby=af cd
dbyaf J J
,

aey\ '
9

and
after the

= af~ cd
ae

fame manner, x

do

4.

EX-

Chap*

12.'

ALGEBRA* EXAMPLE t
x+$y~
8b

8j

Supp.

S$x+yy=zioo
L 3

then

y=

5x80 3x100 =

and #

=
39

6 5x8 3x7

100
19

aJU = 0l9l>
c

I2-|4.

EXAMPLE
3#

II.

=
,2Jf

60

- 3 x 90 _ 640- 270 370 ->4X 2 3x8"" 8 24 _ 3 2~*

4X 1

"^

THEOREM
87. Suppofe

II.

now

that there are three un-

the

known quantities and three equations, unknown quantities x, y, and z.

Thus

then call

m <dx-\-cy-\-fz n
cax-\-by-\-cz
( gX-\-J:y-\-kzz=:p
ae^~' ahn

Then

(hall

z :=

aek

<

+ dhm ~ dbP +gbngem ahf-\- dhc dbk -\-gbf gee

all

' ^

Where

the numerator confifts of

the dif*

ferent produces that can be fnade of three oppofite coefficients

is

taken from the orders in which z

not found

and the denominator

ednfifts

of

ail

the products that can be

made

of the three
oppoliter


84

of

^
the
laft it

Treatise
unknown

Parti.

oppofite coefficients taken from the orders that

involve

three

quantities.

For,

from the

y ' y J
ae

57: ae do ap akz jrmA-gcz "^ 3: _

=
ah

appears, that an dm-4-dcz afz

,
,

and that
,

gb
g

therefore
,

an~ afz*- dm -\- dcz

db

apakz
ah

gm-\-gcz

gb

an

afz dm +
X
ap

dcz

gW*
db

^Wrz ap
akz -f ,Wtf*
&//
-

X & gb X ^ afz + gm akz+gcz X ae gbdcz*

fides,

Take
#//

gbdcz from both

and

di-

vide by 0, fo fhall

afz + dcz X h gbn-\-gbfz = akz +gcz x dbp + dbkz. dp

dm
gib
e

Tranfpofe and divide, fo fhall you find ahn -f- dhm dbp -f- gen -gem a tp ^p,

~"

ack

ahf-\-

dhc dbk-^-gbf
found

gee

lues

of* and j

are

after the

fame manner,

and have the fame denominator. Ex gr* dep -j- gen akn -j- dkm gfm afp
*'

aek

ahf-^dhi- dbk-\-gbfgec*
is

If any term

wanting

in

any of the three

given equations, the values of z and y will be found more fimple. Suppofe, for example, that

will vanifli in the fecond equation,

third,

and k are equal to nothing, then the term fz and kz in the and z
J

j^-^ = ^P-^+Sjn-dip+
gen-dep
dhc

gee

If

Chap*

3.

ALGEBRA.

85
four

If four equations are given, involving

unknown quantities, their values may be found much after the fame manner, by taking all the products that can be made of four oppofite
coefficients,

and always prefixing contrary figns

to thofe that involve the products of

fite

two oppo-

coefficients.

CHAP.
Of
8%.

XIII.

Quadratic

Equations.

TN
X

the folution of any queftion where

you have got an equation that involves

at the

one unknown quantity, but involves

fame

time the fquare of that quantity, and the pro-

dud
tion %

of

it

multiplied by
is

fome known quantity,

called a Quadratic equa-

then you have what

which may be reiblved by the following

RULE
x.

"

Transport all the terms that involve the

quantity
to one fide^

uft*

known

and

the

known

terms to the other fide cf the equation.

2
.

Jf the fquare of the unknown quantity

tiplied by

is

mut*

any

coefficient,

you are

to divide all

tlx terms by that coefficient*

that the coefficient

of

86

A
unit*

Treatise

of

Part

of the fquare of the unknown quantity may be

3.

Add

to both fides

the fquare of half the coef-

ficient prefixed to the

unknown
then

quantity itfelf

and unknown
fquare.

the fide of the equation that involves the

quantity

will

be

a compleat

4. ExtratJ the fquare root

from both

fides

of the

equation

which you will find, on one Jide^ albe the

ways

to

unknown

quantity with half

it
;

the forefaid coefficient fubjoined to

fo that

by tranfpofing this half you

lue

may

obtain the va-

of the

unknown

quantity expreffed in

known

terms"

Thus,
Suppofe y z

+
.

ay

= b,
.

Add

the fauare of

a1

to both fides

3 y
,

4-4
7
.

a*

Extradt the root,

+ r=

. N/^

~
4

Tranfpofe
2p.
-4-

= + J b + %
as
-

aa^

The fquare may be -f- #,

laft

or a

root of any quantity,

3

and hence,

" All
y
z

quadratic equations admit

of two folutions."
finding that

In the
ay

example,
b 4--

after
-

+=

it

may be

inferred that

+ 7 = + J h + -, QXtQ~~J b +

-, fince

Chap. 13.
b

ALGEBRA. J + - X J + 4 4 _
a a

By

-~ gives b

+ ~,
4

as

well as

+4

J
;

C 4

There

are therefore

two values of y

the one gives

= +V5+~

~,

the other

90. Since the fquares of all quantities are pofitive, it is plain that " The fquare root of a negative quantity
affigned."
is

imaginary, and cannot be

Therefore there are fome quadratic

equations that cannot have any folution.

For

example,

Suppofe y z
thenjy
2,

+ 3# == ay %a
ay
z

o,
; 2

add
4

to both,

4
a

= -3^ + - = ~ 4

extract thp root,

-v
+

>

and

y
2

7
more

2:
cannot

whence the two values of y muft be imaginary or

impoflible, becaufe the root of

4
poflibly be affigned.

But of this we Second Fart.

(hall treat

fully in the

Suppofe

gS

^Treatise f

6f

Part

to be refolved

Suppofe that the quadratic equation propofed ayz=zb\ is

then

/ +, ay + T at i + -s 4 4
y

^-iv/* + ?.
2

and j

=|+
in

,/*+.-

If the fquare root of h A M

*2

cannot be extracted

exactly,

you mud,
jy,

order to

determine the

value of

nearly approximate to the value of

\ 4

by the Rules
will

in

Chap,

8.

The
Rule

fol-

lowing examples

illuftrate

the

for

quadratic equations,

EXAMPLE
To
fame number
Call the
',

I.

find that number , which if you multiply by

8, the prGclutt jhall be equal to the fa Hare cf the

having 12 added
\

to it.

number y
J
a

then

tranfp.

Add the

fq.

extracl the root

y of4,j* 87+ 16 = 12 + 163=4, y4= +

%

+12 = 8v,
%y

12,

2,

tranfp.

= 4 4^ 2 ==

6, or 2,

EXAMPLE

Chap.

13.

ALGEBRA. EXAMPLE

Sg

II.

To find a number fuch that if you fubtraft it from 10, and multiply the remainder by the num~
her
itfelf,
it

the produft Jhall give 2

Call

y
is,

then

10
that
tranfp.

iqy

add the

fq.

of 5,

extract

^xjy = 21, yy = iojy= 21, y + 25 = y iqy + 25 = y 5 = + \/ 4 =: i

2
1

4,

2>

and y == 5

= 7?

or 3-

EXAMPLE
?#
b,
y#/;z <?/

III.

two

quantities is #, their produft

Qu. What

are the quantities?

. . .

Cx
Suppofe-J

+y = a
,
y

and #

= a y and* = i,

extract

= y = ; and y ay=z add 4- = + J T4 4 4 - = v * + 7* V,

therefore #
J'

ay
1

tranfp.
. .

b,

ay
i

jy

and

y-^J ^-r^>
EX*

go

Treatise

c/

Fart

I.

EXAMPLE
The fum of two
of their fquares
b.

IV.
a,

quantities

is

and

the

fum

Qu.

the quantities ?

UPP

1x

invol. ^

whence ^ 2
tranfp.

2^y+J
7

+f b = ^ 2y+j%
. .
.

= j\
>

t;::::

C 2y

and

di-

<

vide.

^ "^ i^rs^f,

2ay=ib a\
b
tfj

add -, 4
extr.

+ 7 = - + - = 4 4 4
>

V, y

~ J
a

and^==
2b -^-

Orthuss y=s=~^

+ \/lb

z
,

andtf

-= = a+'Jib-r*a
:

EXAMPLE
$iU amounts to 175 Jhillings
not allowed to fay,
\

V.

^f company dining together in an inn, find their

two of them were

reft

and the
10
s.

found that

their
all

Jhares amounted

to

bad paid.
Suppofe

Qu. How
their

a man more than if many were in company ?

;

number x

then

if all

each man's fhare would have been

>2

17
^

had paid,
,

feeing
It
is

is

the

number of

thofe that pay.

therefore,

by the queftion,
*75

Chap.

13.

ALGEBRA.
175^+350 = 2o#; 20*1=350, x 2^=35; x 2x+ 35+1=36,
io,y
2
z

91

and
that
is,

175.V

io.y*

and
add
1
.
.

extr. v'~~-

=+
1

6,

x
It
is

= + 6 = 7,
1

or

5.

obvious that the pofitive value 7 gives the folution of the queftion ; the negative value

5 being,

in the prefent cafe, ufelefs.

EXAMPLE
and
the
difference

VI.

There are three numbers in continual geometrical

proportion; the [urn of the fir[I and fecond
is is

10,

of the fecond and %d

24.

Qu.

the numbers ?

Let the firft be v, and the fecond will be 10 and the third, 34 x\ therefore,

v,

and 34*
tranfp.

x
z

5 54*
x

anddivid.c

M V x *l^^ 22
cxtrad

2jx = 50,
'

= 100 20* + x =IOO + #*

z

10

10

34

z
-,

#,

T 4 7J9_ + 7J9 =Z _4_ 3

2

W
4
22,

v~. .-*
and
x

- 2 = 7^ = + 4
2.

rr-^-S.
2

or

= ^^ =2*,or2. ?' 2
So

92
So
the

.//Treatise

of

Part

three continued proportionals are

2:8:

32, or
r5
*

9-

2m b\ 9 1 Any equation of this form y -f of the greateft index of the unknown quanwhere
tity

is

term,

may
this

double to the index of y in the other be reduced to a quadratic z z +azz=:b r

by putting y m

= z,

and confequently y zm
as

=z

z
.

And

quadratic refolved

above,

gives

And

feeing j w
*

=z=

+ N/^ + -

EXAMPLE.
The produB of two
quantities

I.
is

a,

and the

Jum of

their fquares b.

Qu.

the quantities ?

whence
mult, by j*
. .

by
%

= f
y*
4

y* = %
y*
..

tranip.

j
1

=
.
.

a*.
4

Put now j;
and
it is

=s

and confequently

~ 2%

Chap.

ALGEBRA. bz = a add-, % bz + - = T 4 4 4
13.
2*
z

f%

tf%

extV ,2
and 2

- = N/- ^,
=+J 2
a
x \

and, kclngyzz Vz

EXAMPLE

II.

2$ //zi a number from the cube of which if you fubtracl 19, and multiply the remainder by
that cube, the producl fhall be 216.

Call the

number
J

required x

and then, by

the queftion,

#
x

19# = 216.
3

i9X#
z=z
.

=2i6,
6

Put

x*
t

= 2%

and
-,
.

it

will

be
1225

361

361

and V~~ .2

22
=
*9

2 c- Hh 35

whence z
But x

~ 3 1 = 27,- or ''
2
i

3=

-}-

8.

= V2

^wherefore

#=

3,

or 2.

94

<A

Treatise

of

Parti.

EXAMPLE
T'o

III.
i

find the value of

8.
3

x 9 fuppofing that x

7^=

Put aF == z, and ^ l then z 7% 8,

= z*
8i

=
.

49

_i = +
2

9, 2

2=8.
But x 1

= z%

and *

= v'z

== v7 ^4

4-

CHAP.
92.

XIV.

Of Surds.

TF a A fo
be an

lefier

quantity meafures a greater

as

to leave no remainder, as 2^
it

meafures io#, being found in

faid to
is

five times,

it

is

aliquot part

of

it,

and the greater

faid to

be a multiple of the
is

lefier.

The

leffer

quantity in this cafe

the great eft common mea;

fure of the two quantities

greater, fo
tity can
it

for as
itfelf,

it

meafures the

alfo
it

meafures
that
is

and no quanitfelf.

meafure

greater than

When

a third quantity meafures any

two proit

pofed quantities, as %a meafures 6a and ioa,

is

Chap.
is

14.

ALGEBRA.
called their greateft

95

faid to
;

be a -common meafure of thefe quantiif

ties

and
it is

both,

no greater quantity meafure them common meafure.

Thofe quantities are faid to be commenfurable which have any common meafure ; but if there can be no quantity found that meafures them
both, they are faid to be incommenfurable

and if , any one quantity be called rational, all others that have any common meafure with it, are alfo
called rational
:

But thofe that have no

it,

common

meafure with
ties.

are called

irrational quanti-

93. If any two quantities a and b have any

common

meafure

,r,

this

quantity x lhall alio

meafure their fum or difference a

be found in a as
;;;,

b.

Let x
in

fo that a ==
is

many times as unit is found mx and in , as many times

;

as

unit

found

in

;/,

fo

that

b~nx\
fo that
is

then fhall

a^bmx+nx = m+nXX',
found
in

fhall

be

a^fby

as often as unit

found

in

mJ^n:
?n

Now

fince

and n are integer numbers,

unit,

If n

muft be an integer number or

fore x

and there-

muft meafure a~+b.

is

x meafure any number as a, it muft meafure any multiple of that number. If it be found in a as many
94. It
alfo evident, that if

times as unit
then
it

is

found

in

m
is

fo that

mx
;

will be

found in any multiple of a 9 as

found in

na 9 na

as

many

times as unit

mn

for

= mnx.
95-

96

'<&

Treatise

of

Part

95. If two quantities a and b are propofed, and b meafure a by the units that are in m (that is, be found in a as many times as unit is found
in

m) and
it

there be a remainder c

and

if

x be

fuppofed to be a
h*

common
>

meafure of a and
c.

mall be alfo a meafure of

For by the
as

fuppofition #

= * ~j~ c
x
is

f ince

ft

contains

many
is

times as there are units in

and there
a

c befides of
r.

remainder

therefore

mb
,

23

Now

fuppofed to meafure a and

and
to

therefore

it

confequently a
r.

mb

meafures
(Art.

mb

(Art.

^)

94J and which is equal

If c meafures b by the units in , and there be a

remainder d, fo that r= nc -{- ^, and

then fhall # alfo meafure d\ becaufe

nc = d
is

it

fup-

pofed to meafure
it

b,

and

it

has been proved that

nc,

meafures

c,

and confequently
is

and b

nc
as

(by Art. 94.) which

the remainder c

equal to

d.

Whence,

after fubtracting b as often as poflible

is

meafured by x

from a y and after

the fame

fubtracting c as often as poflible from b, the re-

mainder d
reafon, if

is

alfo

meafured by x\
as

fo, for

you fubtracl d
:

often as poflible

from
ftill ftill

r,

the remainder (if there be any) muft

be meafured by x

and

if

you proceed,

fubtracYing every remainder from the pre-

ceding remainder,

till you find fome remainder which fubtracled from the preceding leaves no
it,

further remainder, but exa&ly meafures

this
laft

Chap. 14.
laft

ALGEBR A.
ftill

97

remainder will

common

be meafured by #, any meafure of a and b.

96. The laft of thefe remainders, viz. that which exactly meafures the preceding remainder,

muft be
c

common
laft

meafure of a and b

fuppofe that d was this

it

remainder, and that

meafured
(hall

by the

units in r, then mall c

= rd

and we

have thefe equations,

'

a
b

z=i

mb
nc

-\- c>
-J-

J,

Now
muft

it

is

plain that

fince

meafures

c, it

alfo

meafure nc9 and therefore muft meay

fure nc -\it

or

b.

And

fince

it
;

meafures b and
fo that
it

r,

muft meafure mb -\- r, or a

muft be
for

common

meafure of a and

&.

But

further, it
,

muft be
every
d,

their

gnat eft common meafure

;

common

meafure of a and b muft meafure

by the

laft article

and the

greateft

number
is

that meafures J,
greateft

is itfelf,

which therefore
b.

the

common
if,

meafure of a and

97. But

by continually fubtracting every

remainder from the preceding remainder, you

can never find one that meafures that which precedes

will

no quantity can be found that meafure both a and b ; and therefore they
it,

exactly,

will be

incommenfurabk to each other.

For

if there v/as

any
it

common

meafure of thefe

quantities,

as

xy

would

necefiarily

meafure
all


A
y

98
all

Treatise
r, c>
*/,

of

Part
it

L
nc n

meafure a mb
creafe in fuch a
rily

the remainders

&c.

For

would
de-

or

and confequently
thefe remainders

or J; and fo on.

Now

manner, that they

for c
,

will neceffa-

become

at length lefs than Xi or

:

any atfign-

able quantity

caufe

c is lefs

than

muft be lefs than k ; beand therefore lefs than mh9

and confequently lefs than \t-\^\mb 9 or ~a. In for d is lefs like manner d mull be lefs than ~b than r, and confequently lefs than ~d inc9 or
;

4&

The

third remainder, in the

fame manner,
than ia
that every

muft be

lefs

than \c9 which

is

itfelf lefs

thus thefe remainders decrcafe

fo,

one
it

is

lefs

than the half of that which preceded

next but one.

Now

if

from any quantity you

half,

take away more than

its its

and from the

and proceed
remainder
in
lefs

remainder more than

this

half,

manner, you

will

come
r,

at a
It

than any aflignable quantity.

fore that if the remainders

appears there-

d>

&c. never end,

poffibly

they will become

tity,

lefs

than any aflignable quan-

as x>

which therefore cannot

b.

mea-

fure them,

and therefore cannot be a

common

meafure of a and

98. In the fame way, the greateft common meafure of two numbers is difcovered. Unit
is

common

meafure of

all

integer numbers,

and two numbers are

other,
fure than unit
-

faid to

be prime to each

when they have no

7

greater
25.

common meaSuch always

are?

fuch as

9 and

Chap. 14.
are the

ALGEBR A.
;

99
had any

lead numbers that can be aflumed in

for if thefe

any given proportion

common
arife

fure
lefs

meafure, then the quotients that would by dividing them by that common meawould be in the fame proportion, and being than the numbers themfelves, thefe num-

bers

would not be the

y

leaft

in the

fame pro-

portion
99.

againft the

fuppofition.
in

The

leaft

numbers

any proportion

always meafure any other numbers that are in

the fame proportion.

Suppofe a and b to be

the lead of

all

integer
c

numbers

in

the

fame

proportion, and that

in that proportion,

and d are other numbers

c,

then will a meafure

and
and

b meafure

d.

For
d^

if

a and b are not aliquot parts of

then they muft contain the fame

c
c,

number of

the fame kind of parts of

and

d,

and therefore

dividing a into parts of

and b into an equal

number of
the
is
fir ft
;z,

like parts

of

d,

and calling one of

n
%

w, and one of the

fo will the
ail

latter
all
is,

then as

m
b
%

to

fum
j

of

the

ms
:

be to the
: :

fum of

the #'s

that

therefore a and b will not be the leaft in the

fame

proportion

againft the fuppofition.

Therefore
d.

a and b mult be aliquot parts of c and

Hence
for
lefs

we

fee that

numbers which
in

are

prime to each
;

other are the leaft in the fame proportion

if there

were others

the fame proportion

than them, thefe would meafure them by the

fame

ioo

Treatise

of

Part

fame number, which therefore would be their common meafure againft the fuppofition, for

we

fuppofed them to be prime to each other.

ioo* If two numbers a and b are prime to

one another, and a third number

of them
if c
<z,

meafures one
b.

it

will be

prime to the other

For

and b were not prime to each other, they

would have a common meafure, which becaufe it would meafure c, would alfo meafure a, which therefore a and b would have is meafured by
<:,

common

meafure, againft the fuppofition.

101. If two numbers a and b are prime to

r,

then (hall their producl ab be alfo prime to

if

For

you fuppofe them

d,

to have any

common

meafure as

and fuppofe that d meafures ab

by the d a
:

units in e y fo that de
:

ab? then fhali

c,

e.

But

fince

d meafures

and

c is

fuppofed to be prime to <2,

that

it

follows (by Art, ioo.)

,

d and a
d
it
is

are

prime to each other

and there;

fore (by Art. 99.)

'fince

d muft meafure
r,

and yet
is
:

fuppofed to meafure

which

prime
that
is r

to ,

follows that

is

alfo

prime to b
it

d
is

is

prime to a number which

meafures, which

abfurd. 1C2. It follows

from the

laft

article,

that
will

if

a and

are
c
:

prime to each other, then a z

be prime to
quently

to bj then ab will be equal to a z

For by fuppofing that a is equal and confe;

be prime to
a.
c.

a 1 will
c
z

In the fame

manner

will

be prime to

103.

Chap,

4.
If

ALGEBRA.
ih^j Art.

lot

103.

prime
ab

to other

two numbers a and b, are both two c 9 d9 then fhall the product
101.)
c

ab be prime to the product cd\ for

will

be prime to

and

alio to

d 9 and therebe prime a and

Fore,

by the fame

article,

cd

will

to ab.
104.
are

From

this it follows, that

if
1

prime to each other, then

2
,

fhall a

be prime

to
c

<:

by fuppofing,
It
is

in the laft,

that dz=zb^

and
any

d.

alfo evident that

will be

prime

to c\ and in general any

power of a

to

power of
105.

whatfoever.

Any two numbers,

a and b 9 being

given, to find the leail numbers that are in the

fame proportion with them, divide them

'great eft

by their

common meafure x 9 and

b.

the quotients c

and

Jhall be the haft numbers in the

fame proporiri

tion

with a and
if there

For
to be e

could be any other numbers

that proportion lefs than c

and

d,

fuppofe them

and /, and thefe being in the fame proportion as a and b would meafure them: And

number by which they would meafure them, would be greater than x 9 becaufe e and / are fuppofed lefs than c and d9 fo that x would not
the

be the

greater!:

common
.

meafure of a and b9

againft the fuppofition.

106.

Let

it

be required to find the

leaf!

number

that any

two given numbers

as

a and b
to

can meafure.

Firft, if they are

prime

each

eth&i

*2
other,

A
if

Treatise
is

of
the leaft

Part

then their produtl ab

number

which they can both meafure.

For
ab as
c,

they could meafure a

c is

lefs

number than
\

fuppofe that
is lefs

equal to ma, and to nb

and
lefs
lefs

fince c

than ab, therefore

ma
lefs
:

will

be

than ab, and

lefs

than b

and nb being
than a
:
:
,

than ab,
fince

it

follows that n muft be

but

ma nb, and
are

confequently a

:m 9

and a and

prime to each other,

it

would

follow that a would meafure n, and b meafure

m\
lefs,

that

is,

a greater
is

number would meafure a

are not

which
if

abfurd.

But
is

the

numbers a and b

prime to
meafure

each other, and their greateft

x,

common

which meafures a by the

then fhall an (which
:

units in m,

and

meafures b by the units in

n, fo that
is

az=.mx, and

b~nx\
caufe a

equal to bm, be-

mx

nx

n,

and therefore

ail.

= tej
number
a
:

be the

leaft

number
if they

that a and b can

both meafure.
c lefs

For
::

could meafure any

la

than na, fo that c=.

k
:

= ,
is

then

::

ly

and becaufe #

fuppofed
,
it

to be the greateft

common
7;

meafure of a and

follows that
in the

and

are the leaft of all

fame proportion, and therefore

numbers m meafuppofed
than
na,.

fures k,

and n meafures L
lefs

But
is,

as c

is

to be

than na, that

la

lefs

therefore

/ is lefs

than n, fo that a greater would

which is abfurd. Therefore and b cannot meafure any number lefs thaa &
leiler,

meafure a

ndi

Chap. 14na
;

ALGEBRA.
from
this reafoning,
f,

103

which they both meafure, bccaufe na

that if a

wb.
It follows

and

b meafure any quantity

the leail quantity na^

,

which
fure
c.

is

meafured by a and

will alfo

mea-

For

if

you fuppofe
that n
la or r.

as before that c 5= lay

/,

you

will

find

muft meafure

and na

muft meafure

10/. Let a exprefs any integer number, and

any fraction reduced to

its

lowed terms,

fo

that

and n may be prime to each other, and

coniequently an -f

alfo

prime

to

;/,

it

will fol-

low that an
confequently
terms,

+
an
n

will

be prime to

n\ and

^m
2

will be a fraction in its lead

and can never be equal to an integer

Therefore the fquare of the mixt
is

number.
ber a
-\

num-

dill

a mixt number, and never

an integer.

In the fame

manner the cube,

biis

quadrate, or any power of a mixt number,

dill a

mixt number, and never an integer.

muft be

It

follows from this, that the jqitare root of an integer

an integer or an incommenfurable.
is

Suppofe that the integer propoled

that the fquare root of
it is lefs

B,
1,

and
but

than a

-\-

greater than a, then

iurable
;

it

mud

be an incommenlet it

for if

it is

commeniurable,

be

where

reprefents

any fraction reduced

to

IG4
to
its

-^Treatise
lead terms
;

tif

Part
that a A-

it

would follow

fquared would give an integer number B, the

contrary of which
108
It

we have

demonftrated.
laft

follows from the

article,

that

the fquare roots of all numbers but

of

1,

4, 9,

16,

25, 363 49, 64, Si y

100, 121,

144, &c.

(which are the fquares of the integer numbers

i> 2, 3, 4, 5>

6>
;

7^

8 > 9' IO > Ix > 12 >

arables incommenf
#<?
r^i?/j

after

the fame

S?n) *r* manner, /&*

the cubes

#/ 0// numbers but

5, 6,
7,

of

of

i, 2, 3, 4,

8,

9,

C5V. are incommenfuare

rabies

and quantities that

or

to one another
alfo

in the proportion of fuch

numbers muft

have

their fquare roots

cube roots incom-

menfurable.
109.

The

roots of fuch

numbers being

in-

commenfurable
c
2,

are exprefled therefore

by plaj

cing the proper radical fign oyer them

%
5

thus,

V35 v5 V6 ^7? V8, v/l0 5 &V- exprefs Thefe numbers incommenfurable with unit. numbers, though they are incommenfurable themfeives with unit, are commenfurable in power
4/2,

with
is,

it,

becanfe their powers are integers, that

multiples of unit,

They may

alfo be

com-

menfurable fometimes with one another, as the

^8, and the

other as 2 to
rrson

v%
1
:

becaufe they are to one an-

And when

they have a

com-

meafure, as

V2

is

"Jfe

common

meafure
of

Chap.

14.

ALGEBRA.
their ratio
is

105

of both, then
furable

reduced to an ex-

prefiion in the leaft terms, as that

quantities,

of commentheir

by dividing them by
This
in

greated
fure
is

common
found
as

meafure.

common meaquantities,
is

commenfurable

only the root of the

common
divifor.

meafure

to

be made their
V4-

common

Thus

r=

2>

and

?-

=:

^a.
may be
by
reduced

no.

rational quantity

to the

form of any given

furd,
is

raifing the

quantity to the power that

the

denominated by

name of

the furd, and then fetting the rait

dical fign over

5

thus, a

= ^a
^16

== Ja %

*Ja
5

V^j and 4

=:

==

= 34
6*4

^a*

^256

= =

^1024

= ^4".
furds

111. As
ethers of the

may be

confidered as powers
they are

with fractional exponents,

reduced to

fame value that

by

Jhall have the fayne

radical fign^
nents

reducing

thefe

fractional expo-

to fratlions

having the fame value and a

common
771

denominator .
1

Thus
777.

*Ja
77

=
,

n
,

and

Ja

=
n

a m . and

I = = n Tim
I
,

nm

and there-

fore s'a

and

</a,

reduced to the fame radical

fign,

io6
fign,

yf
nm
'

Treatise
nm

of

Parti,

become v m and VV.

t
3

If

you are to

re-

duce
fider
i

s/i

and
as

\]% to the

fame denominator, con-

V3

equal to

j%

the

as equal tQ

T whofe
,

indices reduced to a
x
zzz
6

common

deno6
,

minator, you have %

2,

%
6

and 2 T

=2
3

and
6 *

confequently
6

V3

= v'S

== \/ 2y, and

^2
*

= V2
3

=.

^4

fo that the

propofed furds
6

V3 and

^/2
6

are reduced to other equal furds

^27 and

</4>

having a

common

radical fign.
the. fame

112. Surds of

rational quantity are

multiplied by adding their exponent s %

and dividea

ly fubtrcMlng them.

Thus s/aXy/az^a^xa^ a
and

== aT

= v^

>

--.

a?

==

# = ^-

6
S

V'2

= ^32

jr 2
5

-TV/2

= V 2.
/

113. If the furds are of different rational

H
1
*

quantities, as *
*f

v^ and ^~b\ and have

thefe

the fame
into

fign,

multipy

rational quantities

om

another.

Chap. 14.
another*
fet

ALGEBRA.
divide

107
another*
their
n

or

them by
n

one

and
pro-

the

common

radical Jign

ever
n

du%

or quotient.

Thus Va1
m
sf a*

^=
I d*

*J"aFb^ ;

la*

9 __

/ JL

4/24
If the furds

have not the fame radical fign,

1 1

reduce them by the

ith Art. tofuch as Jh all have

the fame radical fign*

and proceed as

before.
2

Thus
3

i/a

<Jb

= V abm

=
Ja
6

in

rnn

/^-5

I an

<J%

Xy/4
6

*r

1/4

4 4

/7

/IS

= \/ 2.
ents, their

If the furds

have any rational

coeffici-

product or quotient muft be prefixed.

Thus 2V3 X 5i?6= 10^18.

114.

The powers of

furds are found as the

their

powers of other quantities, by multiplying

exponents by the index of the

power

required.

1 hus
J

the

fquare

of

Ji

is

= ^4

tne

cube

io8

-^Treatise

of

Part

5* cube of v/5 == 5 a vi 7 5. Or yos need only, in involving furds, raife the quantity

under the radical fign to the power required^

continuing the

fame radical fign

is

unlefs the index

of that

power

equal to the

name cf
Evolution

the furd,

or a multiple cf
the

it*

and

in

thai cafe the

power of
is

furd becomes

rational.

per-

formed by dividing

the

fraUion which

is

the ex-

ponent of the furd by the name of the root re*

quired.
3
3

6__
z

Thus

the fquare root of

s/ a**
mx

is

\/ a
m

or

y 'a*.
in like

m
15.

The
if a

furd

*/

as/ x\ and

manner,

power of any quantity of the fame

name with

the furd divides the quantity under

the radical fign without a remainder, as here a m divides a mx, and 25 the fquare of 5 divides

y$, the quantity under the fign in s/l 5> with-

out a remainder, then place the root of that

power
tient

rationally before the fign,

fign,

and the quowill

under the

and thus the furd

be reduced to a more fimple expreffion.

Thus

V75
3

==

5^3
J

V48

s/zxi6

4V3

y'8i

s/2jX3
When

== 3V3-

1 1 6. duced to

furds

by the

laft article are re-

their leafl exprefiions,

if

they have

the fame irrational part, they are added or fubtracted 3

Chap.
tracted,

14.

ALGEBRA.
and prefixing the fum or

109

by adding or fubtrafting their rational

difference to

coefficients\

the

common irrational part.

Thus ^75
3 3

*/8i

^54

==: tfy'*

+ V^8 = 5\ 3 + 4/3 = 9V 3 + V24 = 3^3 + 2 V3 = 5V3 V150 = 5v/6 3v/6 = 2^6; s/lfrc + y/J** + V* == ^+ ^ X \/x.
;

>

7.

Compound furds are fuch

as confifl of

two or more joined together. The fimple furds are commenfurable in power, and by being multiplied into

themfelves give at length rational

compound furds multiplied into themfelves commonly give ftill irrational products. But when any compound furd is proquantities; yet

pofed, there
tiplied
<s/a

is it

another compound furd which mulgives a rational producl.

into

s/b multiplied

by */a

Thus

</b gives a

and

the invefiigation of that

furd which multiplied

into the propofed

\s

furd will give a rational producl

made

eafy

by the following Theorems.

THEOREM
118. by a n m

I.

Generally,

if

you multiply am
b zm
n

#*.

a 2m b m

+a

H ~i m

+ a vbi m
b"
:

&c.
to

continued

till

the terms be in

number equal
for

the product lhall be a n

no
X
am

Treatise

of

Parti.

bm
71

a* -\-

- mb m -L- a b zm -f a n -3 m b> m &c. ~ mb m a n "'xmb 2m a n~ 3^3*, &c. an

Z7r,

THEOREM
^ i*

II.

n ~~' lm

b rn -J- a*~~i m b 2m
-{- w >

#*-"

Z?3',

&c.

multiplied by a m
is

gives *

#",

which

demonstrated as the other.

is

Here

the fign

of b n

pofitive,

when

is

an odd number.
is

119.

When
leaft

any Binomial furd

integer

propofed,

fuppofe the index of each number equal to m* and

let

n be the

number that

is

meafured
n
y

hy

then Jh all a n

~m +
1

an

~ 2m b m -{- an v b zm

&c.

give a compound furd,

which multiplied

into the

propofed furd am ~^ om will give a rational produft.

Thus
3

to
3

find

the

furd

which multiplied by

i/a

>Jby will give a rational quantity.

Here
meafball

-f,

and the
-J-

leaft
;

number which
let

is

fured

by

is

unit

i,

then

&n-m

^
+
<z

&r**fc

a"-i m b zm 9

S^ffi
s/ab
gives

+ aF =za^+

= a + fib* + P = s/'a .+
&c.
1

l " T

s/b l , which
h.

multiplied by

^a

ifh

To

Chap. 14.

ALGEBRA.
+

112

To
y/ b*
? == -J

find the furd

which multiplied by -/a?

*, gives a rational product. Here - CT a" 2m b m a n i mb zm9 and == 3 and

a*

&c.

=a

"~

T --

T^ +

0* __ a^p

j^ a *fri

^T = = </a \/a ^ +
+
3

tf

fr+

THEOREM
m 120. Let a

III.
^JT

a n-2mfri
product

_}.

an

+ -^ nb +
zl

~ be multiplied by a n n
a n -V"fc l
:

+ &V.

and the

fhall

give a n +_b m

therefore n mnjl be
alfo

taken the Uaft integer that Jhall give

integer.

an

Dem.

an

-m

T a ~ 2mb + a-l mb ~ a -4 m
n
l

ll

bi !

Xan + V
an

&c.

ab"~

lX

+ <*# +
an

-f-

a n * m b zl, &c.
zm b 2

~ mb
*

a"
*

^ &c.

+ i*
-j-

/7

The
am

fign of w

is

pofitive only

when

is

an odd number, and the binomial propofed

-j-

is

$ 12*.

ii2

121.

Treatise

of
is

Parti,
propofed
Jet

If any binomial

furd

whofe two numbers have different indices, thefe be m and /, and take n equal to the leaft
teger

in,

number

that

is
l

meafured by
a n -i m b* l
furd,
fhall
2,
-

m
n

and by

and - w

+ aH- 2mb +
compound
g

mall give a

&c. which multiplied

~ a -* mfo\

by the propofed a m +. V
duct.

give a rational pro-

Thus s/ a
*.,

%/
-y
n

b being given, fuppofe

tzz?-'4j

and

== -J, therefore you have

l

w=3,

+ a ~ 2mb + a-^b + a -4 mbi 3+ &c. = a + a^F + a^'h* + a ""^ + ^%f + p* = a* + a + *V + ab + 3__ , <^T + s/^ + ^X^ + ^Xv/^'
and 0"-*
"*"

zI

<7 y

4-

S_ 3

7.

3
5

J^ah-\- /a

X \/^ +

\/b =za z ^a + a z X\/b

3 3
'

z 4- as/ax \/b

+ ab + b*/a X */b + bx s/ b
by
*.

which
a*

multiplied
nl

the

s/a

j/b 9

gives

b m 5=

tf

122. By thefe Theorems any binomial furd whatfoever being given, you may find a furd

which multiplied by

it

mail give a rational pro-

dud.
Suppofe that a binomial furd was to be di2,

vided by another, as y'zo

+ Vi2, by </5 -y^*

the

2.

2.

2,

Chap. 14.
the quotient

ALGEBRA.
may
be exprefTed by

113

i_*L

But it may be exprefTed in a more fimple form by multiplying both numerator and denominator by
that furd which
multiplied into
the denominator

gives a rational produffi.

A /20

Thus

s/S

+
5

**

" z=

+ ^I2 j/5 + \/3 _ ^5 + ^3 VS-VS 16 + 2^/60

"
.

v^ioo

s/l 2y/6o

3
is

+6

123.

In general,

when any quantity

di-

vided by a binomial furd, as a m _ b 1, where m and / reprefent any fractions whatfoever, take n
the leaft integer number that
is

meafured by

and

-y, multiply both

numerator
n

md

denominator by
the denomi-

an - m

+a

n ~ 2m l b

+a

: m b" 1

&c. and

nator of the produfi will become rational^ and equal

nl

to

bm

then divide all the members of the

numerator by this rational quantity ^ and the quote

arifing will be that

of the p'opofed quantity divided

its

by the binomial furd, expreffed in

Thus
K /6

= HIV? = V5 + 1
vi
v
2
3
3

leaft

terms.

V2

4/4.2^-^/1%

*/20

^20

^3

^16 + 2 + ^/4
3

__
3

</20
3

2^/2
<^
3

+ 2+^/4
3

v/ I 6

+ 2 + V '4
/

1/4

-\/ 2

2^/2+2+^4

114
s-

Treatise

6f
.

Part
I

2/40+2/20 + ^80 v + -r v Tv
*

2 /5

+ /20 + /io:
/ss4, n

Alfo

T
3

= (becaufe = i,
*

3,

and ^*

9 =
<

1)

4</2o4-4N/ioX^3-^ 2 v/2 X^94-6A/io>f-3 /2ox^3 + WioX'/

= 8/5 4/10
6/10

x /3
3

8/5 x /9
3

re-

6x </$X
When
it

\/3

3/10 x /9is

124.
quired,
root of

the fquare root of a furd

may be found

nearly by extracting the

rational quantity that approximates to its to find the fquare root of 3

value.

Thus

+ 2v/2,
to

we

firfl

calculate

3 +

2/2

= 5,

1, 41421, and 82842, whofe root is found

/2 =

therefore

be

nearly 2,
2,

4T421

fo that

^3 + 2/2

is

nearly
to
-

4 1 42 1.

But fometimes we may be able

exprefs the roots of furds exactly by other furds

as in this
is 1

example the fquare root of

for 1

+ = 3 + 2/2.
2

/2,

a/2

+ 2/2 x + V^2 = + 2/2

3
1
1

In order to
found,
furd,
Jet

know when and how

this

may be
:

us fuppofe that
will

x+y

is

a binomial

whofe fquare
are quadratic

be

x*+y z +
then x z
j

2xy

If

and y

furds,

+- y %
that

will

be rational,

and ixy

irrational

fo

2xy
(hail-

Chap. 14.
fhall"

ALGEBRA.
lefs
z

115
,

always be
is

than x

difference

+y

2#y == xy*,
z

+y

becaufe the

which

is

always pofitive.

Suppofe that a propofed furd

confiding of a rational part A, and an irrational

part B, coincides with this, then x

+y =A
2

and

yy=
2

ife

Therefore by what was faid of

"*

Equations, Ci^p. 13th,

jy

R = A x = j, and B Av and* Ax
z

therefore
-

#*

z
-\

=
B2
4

0;

from whence we have # 2 ==

A
2

2
is

and

a/ a

2-

m
.

z=:

Therefore when a quantity

partly rational and partly irrational

propofed to

have

its

root extracted, call the rational part

A,

the irrational B,

and the fquare of the great eft

member of the J
fquare of the

root J fhall be

2
p^

and the
B*
2

\/

A"

lejfer

part Jhall be

And

as often as the fquare root

of

B
is
2

can

be extracted, the fquare root of the propofed

binomial furd
mial furd.
pofed, then

may

be expreiTed
if
3

itfelf as

a bino-

For example,

4-2^/2

A-3, B~2>/2
2

and

B
=2

pro==

y
z

= herefore# = A \/A* B _ 5=
8
1.
1
2

and

1 heref.

+ y z=n+\/z,

To

n6
To
po&A's=i
:

yf

Treatise

of
i

Parti.
fup-

find the fquare root of

+ V 8,
A
2

i,B = V^-8
^

=
is
1

1
-

22
=

fothat

9,

2, therefore the root required

+ V 2.
y
are not quadratic
if

125. But though x and

furds or toots of integers,

they are the roots

of

like
*J

furds,

as if they

are equal to V'm</z

and

ny/Z) where
s/ z

and n are integers, then

A = m 4- n x
m

ancl

^ = ^ ^2
A

A
2

-B*
:

n\

X
2

z,

and ** ==

4- \/A

'2
+
V -/zfrom
itfelf

== i/z and #
part
its

+j

ffz/2

here eafily diftinguifhes

The B by

being: greater.

126. If x and
-

v are

equal to

V ms/z and
-J-

vzzv7, then #*4- 2 xD'

+ J^ ^^ + ^
2;

^mrVzt.
fures

So that

if

or

be not multiples

one of the other, or of feme number that mea-

them both by

a fquare

number, then

will

A itfelf be a binomial. 127. Let x + y +

furd,
its
2 fquare ~
2

z exprefs any trinomial

+jy

+ + 2xy + 2*2; + 2yz

a

may

be Hippofed equal to

-J-

as before.

But

rather

Chap. 14.

ALGEBRA.
two
radicals as 2xy
z

117
by ixz 9

rather multiply any

and divide by the third 2yz, which gives the

quotient 2x
rational,

and double the fquare of

the furd x required.

there
are

The fame
z

rule ferves

when
s
z

four quantities,

x +jy 2
9

+ +
2

2:

-f-

ixy

+ 2x5 + 2xz + 2yz + 2ys + 2zs

2xs,
Z

multiply 2xy

by

and the producl ^x z sy divided by 2sy

a rational quotient, half the fquare

gives 2X

z manner 2xy x 2yz 4y xz, which divided by 2xz another member gives 2y z 9 a ra-

of 2x.

In like

tional quote, the half

of the fquare of
s

2y.
;

In

the fame
their

manner z and
-\-y
-\-

may be found
1

and

fum x
2yz

z -f- s9
J

the fquare root of the

-fs

feptinomial x z

-j-jy

-{-

zz

+ ixy
try

-{-

2aj

-[.

2xz

-f-

-[ 2jj,

difcovered.
to
find
;

For example,
,0

the
I

fquare

root of

+ v/24 + V40 + V60 which find to be 6 = 4,

I

^~/^,
iX*/8 ==V'2
,

\/

the half of the fquare

root of the double of which, viz.

is

one member of the fquare root required

next

6,

the half of the fquare root of

is

the double of which

^3* another member of

the root required

-,

laflly,

= 10, which
+ v^3
multi-

gives s/ 5 for the third

quired

member of the root reFrom which we conclude that the fquare

10

root of

+ 1/24 +
I

1/4.0

-f-

60
it

is

s/2

+ \/$

and trying you

2

rind

iucceeds, fince

n8
drinomial.

^Treatise
itfelf it

5/'

Parti.

multiplied by

gives the propofed qua-

127. For extracting the higher roots of a

binomial,
are

whofe two members being fquared

is

commenfurable numbers, there

the

fol-

lowing

RULE.
*

" Let
root

the quantity be

A
c

^_ B,
the
leafi

whereof

is

the greater part 9

required.

and
by

exponent of the

Seek the

power n c
iient being

is divifible

AA

number n whofe

BB,

the quo-

Q^

Compute

v A+B x ^Q. in

the
r.

near eft integer number, which fuppofe to be

Divide

AVQ^by

its

great eft rational divifor>

and

let

the quotient be

s,

and
t,

let

in the

neareft integer number, be

required be

fo Jhall the root

the
c

if

root

of

-5-

can be extracted.

Thus to find we have A 2

7,
7,

EXAMPLE of ^968 cube B = 343, whofe

I.

the
z

root

-f-

25,
are

divifors

7,

whence

72

7,
is,

and

Qj=i.
is

Further,
little

A~r B X ^Q,

that

V^68 -f 2 5

more
than

* Arithm. Univerfal. p. 59.

Chap.' 14-

ALGEBRA.
is

119
4.

than 56, whofe nearefl cube root

fore r

Whereits

= 4.

Again, dividing

V968

by

great-

eft rational divifor,

we have AVQj=: 22^2, and

the radical part </2

j,

and
Is

or

~
2^/2

in

the
ft

nearefl

integers,

is

=
and

/.

And

laftly,

= 2V2,

*Jfs z

n =
V968

i,

VQ^= V i

i.

Whence 2\/2+
trial,

1 is

the root, whofe cube,

upon

find to be

+ 25.
II.
>

EXAMPLE
To
find the cube root
z

have A 2

B = 250, whofe
;;

of 68 ^4374

we
and

Thence

divifors are 5, 5, 5, 2.

io,

and

Q^=
2

4j

\/a + B x ^Q,
ly 7

or

^68 + V4374 X
or

is

near-

=r
is,

-,

again,

AVQ,

68x^4 136x^13

that

'+7
1,

7+y
,

10

and
/j

or
*//

-,

is

nearly

=4=
and
tried
is

/.

Therefore
6
3

= 4,

V n =2 </6

v vQ^- ^4 =
V
.

\/2,

whence the root to be

v/2

EXAMPLE
Suppofe the
fifth
z

III.

root of 29^/6
3,

demanded,

+ 41-/ 3

is

and ^==35

(^=81,
'==5

120

-^Treatise
2

of

Parti.

IO

and >/Q^r=: y'Si == ^9.

to be

And
.

therefore trial

made with

"7"

In thefe operations,
tion, or if its parts

if

the quantity

is

a frac-

have a

common

d.ivifor,

you

are to extract the root of the numerator and de-

nominator
to
this

or of the facers feparately.

the
to

Thus
12-,
is

extract

cube
a

root

of

V242

reduced
-*

common

denominator

And

the roots of the numerator

feparately
if

and denominator,
root

found,

give

the

-.

And

you feek any root of

*/2
6

V3993

-f

V17578125,
3

divide

its

parts

by the

common
1 1

divifor

^3, and

the quotient being

+V
1 1

25, the root of the quantity propofed

3

will be

found by taking the roots of s/ $ and

of

+ *J 125,

and multiplying them into each

of

other.

128.

The ground

this

Rule may be ex-

plained from the following

THEOREM.
Let the Jum or
difference

of two quantities x
is c>

and y

be raijed to a

fewer whofe exponent

and

Chap.
and
let

14.
the

ALGEBRA.
ift,

121

3 d,
into

$th, yth, &c. terms of that

power,

collected

one

fum 9

be called

ths reft of the terms, in the even places, call

the difference of the fquares of

A, and B

and B jhall be x and y

equal to the difference of the fquares of

raifed to the fayne

power

c.

For the terms

&c.) are

in the c

power of x

-{-

y
i,

(writc,

ing for their coefficients, refpectively,

e,

d,

x c -\-cx c - Jy-{- dx c

~Y -f ex'-y +,&c. = A + B,

and the fame power of x

figns in

(changing the

the even places)

cx c

~ y -j- d\<l

and therefore

y x+jyf X x
2
l

is
f

<f*

-3)'*-f ,&x.=:

A-B,

y\

=A-f BxA B
x~^y\
)

= A
z

x-i-y

= x y
z

Q.E. D.
Let one, or both, of the quantities
a quadratic furd
,

x, y,
c

be

that

is,

let

x ~\-y, the

root

of the propcfed binomial

of thefe forms,

p -f Wq,

A + B, Wp
-\-

belong to one
or ks/p
-j-

fr

Wq.

And
1

it

follows,

If

x-\-y=p -fnumber
will
;

li/q, that, c

being any whole

will

number, A, the fum of the odd terms,

a rational
in the even places,

be

and B, the fum of the terms

multi-

each of which involves an odd

be a rational number

power of y
2.

plied into the quadratic furd Jq.

Let

c,

the exponent of the root fought,

be an odd number, as we
I

may

always fuppofe
it,

122
it,

A
if it is
;

Treatise
even,
it

of

Parti.

becaufe

may
root,

be halved by the
till

extraction of the fquare

it

becomes

odd
3.

and

let

+ y = ks/p +
B

q.

Then

vv ill

involve the furd */p 9 and

will be rational.

But

if

both members of the root are irra-

aal (x
irrational,

+ y = ks/p + V^j A
all thefe cafes,
it
is

and B are both

the one involving s/p, and the other

the furd i/q.

And
when x
than B.

in
is

eafily

feen that

greater

than y 9

will

be

greater

129. From this competition of the binoA -|- B, we are led to its refolution, as in tke foregoing rule, by thefe fteps.
mial

L
"When
fect c
1.

is

rational^

and

A
z

is

a per-

power.

By
,

iht 'Theorem^

B = ^

jP|

accu-

rately

A*
of

W
r,

and therefore extracting the

it

will be

y\

root of

Call this root n.

c

2.

Extract in the nearefl integer, the

root

A+
3.

B,

it

will

be (nearly) x

j.

Which

put .=

Divide x
is

f- (1=

quotient
divifor
.

(nearly) x
is

;/J> ,

by #

-f- 3/

(=
\

r) the

and the fum of the

that
it

and quotient

(more nearly) ix
is

is,

if

an integer value of x
r

to be found,

will

+~
-.

be the neareil to

2
4.

Chap.

14.

ALGEBRA.
;

123

4.

** ** jy*==y*

or,

+
r
,

=f

whence

jy

==

V ~
/
,

r+ 7i
,

and therefore, put-

ting *

the root fought #

exprefiion as

+y= /+
in the rule,

V/ 2 # ; the fame i, when Qj^z 1, s

that

is,

when A*

is

a perfect
is

power, and the greater

member

rational.
II.

When
By
j

is

irrational,

and

Qj=

1.

the fame procefs, x

w.

(= T)

and

= Vt*
;

onal, and c an

But feeing A is fuppofed irratiodd number, # will be irrational

will

likewife

and they

both involve the fame

j,

irreducible furd /p, or

which

is

found by

dividing

A by
Vtts*

its

greateft rational divifor.

Write

therefore for x or

T,
.

its

value

ts

J5

and #

+J

III.

If the c root of

multiply

A*

2,

B* cannot
(leaft)

be taken,
as that

by a number Q^, fuch

be the
a

the product
n<

may

perfect c

power
of

(= A*Q^ B QJ

And now

(inftead

A+


124

^Treatise
extract the c root of

of

Parti.

A + B)
r

which, found as above,- will be

A + b x ^Q^ ts + V/V n
;

and confequently the

ts-\-*ff ?'
that
is,
.

root of

ts-X- %/t*s
I

n divided
y

A
c

-j-

will

be

by the

root of

^Qj,

ql
It is required in the rule that
c

a per feci

power

(n j be found which Jhall be a multiple of B* by the whole number Q^ To find this power,
7,

let

the given

number
...

A
by

be reprefented
fingle
.

by the product aHfdf,

let

whole
by by
.

divifors

be #,

<z,

*z,

d9

f
a

and

the product of thefe divifors raifed to the power

cy

which

is

m a cb c dc fy divided by a bpdf will give

&rr*lfi~-tfa-*f**

the

quotient

Q^

whole

number, provided fome index,

not greater than
fingle
till
c.

as

or

p9

be

If it is,
z

take, inftead of the

di-vifor

a or by a

or #*, a 1 or b>,

&c.

there be
is,
till

no negative index

in the quotient

that

Q^

be a whole number.

130.

We

may add

the following remarks.

is

1. If the refidual

given,

it

is

evi-

dent from its fame rule gives

2.

genefis
its

by involution,

that

the

root x

j.

The

extracting the c root of

A+

B, or of

A + Bx

^Q^j

in the neareft integer, neglect-

ing the fractional part, will always give x-\-y

fuch, that the value of x which refults in the

operation fhall not differ from

its

true value

by

unity $

Chap.
unity
j

14.
that

ALGEBRA.
is,
it

125

fhall

be the true integer va-

lue fought.

For,

being fome proper fraction,

c

let

+
it

+ / be
and
x
let

the accurate value of

the quotient of x
g->
1

y T
g

y*
if

%/A + B x
divided by

>/Q 5
be

tnen > the

fum of

the divifor and

quotient being 2x
fractional part

_f + g, could make
**,

our reckoning the

in the value of

it

or

/ =

a difference of unity would follow that / g

2.

Which

is

abfurd, g> as well as

/, being a proper fraction. and B are irrational 3. If both

or, if the

lefTer

of the two members

is

rational,

no root
rational,
it is

denominated by an even number can be found.

4.

When

the

greater
c is

member

is

and
is

the exponent

an even number,

am-

biguous whether the greater

rational or furd.
is

member of

the root

And
or,

though a root in the

that failing, in the

form of p -f- Wq the form of k</p

not found, yet a root in

-|- q,

form k\/p
If

+ l*Jq

may

be obtained.
q-,

we look

to fubtract x

for a root k*/p -f

v

we

are

now

from x+y> and half the remain-

der will give y (or q) the rational part. And to (~ n ) adding jy% the fum will be x z x*

F~
So
that y

=
2

,
'

and x

=v
as

-\-n-,

the

'

exprefllons being the

fame

when

is

odd,
with

126

A
no

Treatise
number

of

Part

with the fign of n changed.

fucceed, and a prime
radical fign,

If this docs not

ftands under the

farther trial need be

made.

But

if

a compofite

number

{rands under the

radical fign, tht root

may
%

poiTibly belong to

the form ks/p ber being p

the

+ Wq
k,
/,

and that compofite num-

%<>

fince krp

l q =z ,

and k*/p

= x,

numbers

neared integers,
as in this

may be fought for in the and trial made with Wp-\-ls/q\

EXAMPLE.
^to find the fourth root

of 49849
ft

2 895

V 2 24.
z
9

The

4th root of
of

and the 4th root

nearly:

A B, that and = -^ = 17 nearly.

is,
l
i

A
a

is

157

= x*y =zn x y = r =

()

Whence

= 9~^ = i2.
B
being

But now the

leaft radical fac-

tor in

V14

=V 7
/

x
/*
j

2, I

put 13 (--=x)

== k\Sy, and k in the near eft integer

gain, k*p
is,
/*

root

n= 175 and = x = 5^/7 3^/2.

l*q

=5. A-

18,

2 157 ; that which gives the

manner the even roots may be fought But to avoid ambiguity and needImmediately. lefs trouble, it is better firft to deprefs them by
In
this

extracting the fquare root, as in 124.

A SUP-

Chap.

14.

ALGEBRA.

127

SUPPLEMENT
TO THIS

CHAPTER.
131.

ry^HERE occur fometimes, efpecialiy

A + BV q,
To
si

in the reiblution of cubic equations JL by Cardan's Rule (Part II. 79.) binomials of

this

form

whofe cube roots muft

cannot

be found.

thefe the foregoing rule

be applied throughout, becaufe of the imaginary factor

q.

Yet

if

the root
firft

is

exprefli-

ble in rational
rule
will-

numbers, the
lead us

flep of that

often

to

it

in

fhort

way,

not merely tentative, the

to

trials

being confined

known

limits.

For it being,

univerfally,

and, in the prefent cafe,

vA * .= x*y* VA +B^ (t x* f)
1
2

z=p ~L T x
radical fign

',

if
its

we

divide the part under the

by

greatefh rational divifor, the

quote

is

the imaginary furd

V
/*

q9

and from

vA^B^,
divifor of

fubtracting

p\
/

the fquare of
is

fome

A, the remainder

q,

known
That

multiple of the fquare of

a divifor of B,

128

A
z

Treatise
divifors of
for
z

of

Parti.
and B refpecl*J
z

That p and

/ are
-,

tively is evident

cubing p

+
3p

q,

you

find

A~pXp 3l
/

q,

B
z

the figns of p and

muft be fuch
3l q,
I

produces of

z p xp

And X 3p Pq of the
x
l q.

as will give the

z

fame

figns as

and

refpeclively.

EXAMPLE.
To
8
1

find the

cube root of 8

-j-

V 2700 ==
-

30\/^3.
3

Here A=8i,B=30, = 3; ^81x81 + 2700 =21 == 2 + lz q. Subtracting therefore from 2 1, the fquare of (p) + 3, which is a divifor of A,
there remains (l
is
z

X q ) 2 X 2 X 3.
Laftly,
factor p*

And
z

f/==) 2

a divifor of 30.
pofitive,

A (pxp
;

being
tive,

and the

3/^ negaand for the

is

j/*#J

p mufl have

the negative fign

2.

like reafon / =:

So

that the root

+
as

2\/3T^".
It will

be fhewn in the Second Part of

this

Treatife that

" every cube or other power has

and imaginary,
as there are
;*'

many

roots, real

units in the exponent of the

larly,

power

particu1,

that

unity

itfelf

has the

cube roots

T"

"t 2

ZL2.

T "~"

and

^-^.
3

If therefore
roots, in this
a
3

we would

find the other

two cube
and

example, feeing

z'^^Xij

V2 X ^

=2


Chap.
l

4.

ALGEBRA,
we
are to multiply

129

(z reprefenting any cube whatever, and z any

of

its

roots)

the root already found,

4 iV 3, and the and + 4V-3 will be the root required.

by
4. 4-

W and products 4vC~^

by

+ 2V
3,

3.,

i+

Or, becaufe the denominator of the imaginary roots of unity

half of a divifor of
is

2,

taking
2
1

p zs
2

A, we have
==
4-,

V x3=
as
z

%p

/ #,
z

that

is /

and p*
both
-J

^I q as well
Again
fo the

-J V

.1,

one

q being negative,
is

p and /mud be

negative, and the root

take p=~i, and you

4:^ 3 fhall find = -4-4-5

/

remaining root

is

4- + 4V

We may

3, as before.

here obferve that the operation ought

it

to be abridged, where

can be done, by divid-

ing the given binomial by the greater! cube

that
it

contains
;

and rinding the root of the

will give the root

quotient

which multiplied by the root of the

cube by which you divided,

required.

Thus,

in

the

foregoing Example,
1

81+v^- 2700 =27x3 + ^

roots of 3-j-V

0/0
a 7
>

and the
eafily,

iff being now, more

,

found to be

+ 2V

4-,

4-
3,

w-

-4-,

and

i+W

1-4-,

thefe multiplied

by

the cube root

of 27, give the roots required the fame as above. " If the coefficient of the imaginary member of the binomial has a contrary
roots will
fign,

the

be the fame, with the figns of the

imaginary

ijo

^f

Treatise,
or. 8

&e.

Parti

imaginary parts changed."

Thus

the cube roots

3,

of 8

3
2s/

2700,
si

3>

+ i^^*
l

30V
and

will

be

I |V~^

And therefore s/% 1 + V- 2700 -f n/8 i-vC^oo

= -,3x2 = - 6, or = --|-x2=:-.3, or = |-x2 = 9,

the imaginary parts vanifhing by the contrariety

of

their figns.

We
not,

may

obferve likewife, that

in

fuch

roots,

whether exprefiible

rational

numbers,

or

may be found by
q

evolving the binomial

in pag. 4.1,

A-J-lW fumming
going
87|

by the 'Theorem

and

the alternate terms.

As,

in the fore-

example,

81

30^

31%

or

rather

X 1+44^ 3}%
fum of
the
-J,

feries, the

being expanded into a odd terms will continually

approach to 4.5 *=
efficients

and the fum of the co-J-j

of the even terms to

which

is

the

coefficient of the imaginary part.

But for a ge-

neral

and elegant

folution, recourfe

mult be had
in Philof.

to

Mr.

de Moivre's

Appendix

to Dr. Saunderfon's
it

Algebra, and the continuation of

Tranf.

N45i.

What

has been explained above

may

ferve, for the prefent, to give the

Learner

fome notion of the compofition and

thofe cubes
prifed to
\

refolution

of

that he need not hereafter be furreal quantities

meet with expreiTicns of

which involve imaginary roots.

End

of the First

Part.

TREATISE

ALGEBRA.
PART
Of the
the
II.

Genefis and Refolution of

all

E-

quations of
different

degrees

and of

Affe&ions of the

Roots.

CHAP.
Of the
in general
;

I.

Genefis and Refolution of Equations

and the number of roots an

equation of any degree

may

have.
as the

i.))(3DJ(FTER
fn

the fame

manner

(JT

higher powers are produced by the

multiplication of the lower powers

)*QfiD!(

of the fame root

fuperior orders are generated

-,

equations of

by

the multiplica-

tion of equations of inferior orders involving the

fame unknown quantity.

And

an equation of

any

132

^Treatise
many

of

PartIL

any dimenfions may be conjidercd as produced by the

multiplication of as

fimple equations

as

it

has dimenfions
foever^ if the

,.

or of any other equations whattheir dimenfions is equal to

fum of

the dimenfion of that

equation.

Thus any
as

cubic

equation

may

be conceived

generated

by

the multiplication of three fimple equations, or

of one quadratic and one fimple equation. biquadratic as generated by the multiplication.
of four fimple equations,
equations
equation.
\

or

of

two quadratic

or laftly, of one cubic and one fimple:

2. If the equations which you fuppofe multiplied

by one another are

the fame,
elfe

then the

equation generated will be nothing

but feme
is

power of thofe equations, and the operation

merely involution
ready
:

of which we have treated alis

and,

when any fuch equation

is

given,,
it

the fimple equation by whofe multiplication

is

produced

found by

evolution^ or the extrac-

tion of a root.

But when the equations that are fuppofed to

be multiplied by each other are
other
equations
different ^

then,
\.

than

powers are generated

a different operation
is

which to refolve into the fimple equations whence

they are generated,
involution, and
is
is

from

what

called, 'The refolution

of equations.

But

as

evolution

is

performed by obferving
-,

and tracing back the

fleps of involution

fo to

difeover.

Chap.

i.

ALGEBRA.
carefully obferve

133

di (cover
tions,
tion.

the rules for the refolution of equa-*

their genera*

we muit

and
a->

unknown quantity to be xi 3. Suppofe the any fimple equations to be its values in

b9
c,

d9 &c. then thofe fimple equations, by bringing, all the terms to one fide, become # a o, x 0, x 0, &c. And, the x

=
as

product of any two of thefe,

axx bo
three

will give a quadratic equation,

or an equation of

two dimenfions.
them,
as

The product of any

of
a

x axx bxx c-=.o

will give

cubic equation, or

one of three dimenfions.

The

product of any four of them

dratic equation, or

will give a biqua*

one of four dimenfions, as

axx bxx cxx J-=cO<

"
unknown

And,

in ge-

neral,

In the equation produced, the high eft diquantity will be equal to

menfion of the

the number of fimple equations that are multiplied

by each other

^
any equation equivalent to this
b

4.

When

biquadratic x

axx
9

X x d ~
x

is

propofed to be refolved, the whole

x =:o, x c=:o x d=i
tiplication
it

difficulty con-'

fids in finding the fimple equations

a=zo

o,

by whofe muleach of thefe

is

produced

-,

for

fimple equations gives one of the values of x y

and one folution of the propofed equation. For,

if

134
if

Treatise

of

PartIL

any of the values of x deduced from thofe

the terms of the whole

fimple equations be fubftituted in the propofed

equation, in place of #, then
that
all

equation will vanifh,

and

be

found equal to nothing.

iuppofed that x
then
the

Becaufe
,

when
r,

it-is

= a,

or x

or x =r

or x

=d

product x

axx bxx cxx d

is

does vanifh, becaufe one of the factors

to nothing.

equal

There

are therefore four fuppofi-

tions that give x

a x x bxx cxx d~ o
;

according to the propofed equation

that

is,

there are four roots of the propofed equation.

And

after the

fame manner, "

Ay

other
as

equa-

tion admits of as

many
as

folutions

there are

fimple equations multiplied by one another that

produce

it,"

or

"

many

as there are units in

the highefl dimenfion of the

in the propofed equation."

unknown

quantity

5.

But

as there are

no other quantities what-

foever befides thefe four (a,b,c,d) that fubftitut-

ed

in the

product x

ax x bxx cxx d
make
x
the product vanifh

in the place

of

x,

will

therefore, the equation

aXx bxx cX
more
b,

d = o,

cannot poiiibly have more than thefe

folutions

four roots, and cannot admit of

than four.

If

you
e9

fubftitute in that

product a
nor
c,

quantity neither equal to a, nor

d y which fuppofe

then fmce neither

is

nor
9

eb>

r,

nor

c d

equal to nothing-, their

product

Chap.
product

i.

ALGEBRA.
bxe cxe d
is

135
cannot be

eay^e

equal to nothing, but muft be fome real pro-

duce
tide

and therefore there

no fuppofition be-

one of the forefaid four, that gives a jufl

that

value of % according to the propofed equation.

So
that
it

it

can have no more than

after the

thefe four
it

roots.

And

fame manner
the

appear?,

u No

equation

can

have more roots than

contains

dimenfions

of

unknown

quan-

at,"
6.

To make
in
.v

all
;

this dill plainer

by an ex-

ample,

numbers
+

fuppofe the equation to be

*,3

refoived to be

i r

+ 35#

50*
Xx

-J-

24

= 0,
the

and that you difcover that

fame with the product of x

this
1

equation

is

Xx

and

4, then

you
for

certainly infer that the four

values of x are 1, 2, 3,

numbers placed
confequently x 4

4 x makes
5

feeing any of thefe

that product,

io^

ZS X%

50^ + 24, equal

be no other
:

to nothing, according to the propofed equation.

And

it

is

certain

that there can

values of x befides thefe four

fubftitute any other
tors

fince

when you

i,

number
x

2,

3,

for

x in thofe fac4,

none of the
product can-

factors vanifh,

and therefore

their

not be equal to nothing according to the equation.

7. It

may be

ufeful

fometimes to confider

equations as generated from others of an infe-

rior


136
rior
fort
.//

Treatise
fimple
ones.

<?/"

Part
a

II.

beiide

Thus

cubic

equation

may

be conceiyed as generated from

l
$c

the quadratic

px-{-q
,q /
\

o,

and the fimple

\

equation x

a o,
q
-

multiplied by each other

whofe product

ax

px

-f-

-*p

apx

_Qm

eX r re (*s an y cubic p j

equation whofe roots are the quantity (a) the

value of x in the fimple equation, and the two
roots of the quadratic equation, viz.

--

and I.

-.

44 as appears

from

2
as thefe roots

C&<3#. 13.

Part

are real or

And, according impojjible, two of the

roots of the

cubic equation are real or

8
that
.

impojfible.

In the doctrine of involution

we fhewed

"

the
is

fquare

of any quantity pofitive or

negative,

always pofitive," and therefore " the

is

fquare root of a negative

impofiible or ima-

ginary,"

For example,
but

the

v'a z

is

either

or

a\

can neither be

+^
and

nor

- #,

but mull be imaginary.

flood that

"

Hence is underquadratic equation may have no

its coefficients,

-impofiible expreflion in

yet,

when

it

is

refolved into the

it,

fimple equations
involve impofiible

that produce

they

may
the

expreffions."
x"-\- a*

Thus

quadratic

equation

=o
-f-

has no impofiible coefficient, but

it is
*t-

the fimple equations from which

viz,

produced,
2= o,

da

7
-

=z

o%

and x

a"

both

Chap.

i.

ALGEBRA,
an
is

137
;

fquare a~
root
is

both involve

imaginary quantity

as

the

a real quantity, but

its

fquare

imaginary.

After the fame manner a

biquadratic equation,

when

refolved,

four fimple equations, each of which

may may

give

give

an impomble value for the root

and the fame

that

may

be faid of any equation that can be pro*

;

duced from quadratic equations only

is,

whole dimenfions are of the even numbers.

9.

But

cc

a cubic equation (which cannot

be generated from quadratic equations

only,

but requires one fimple equation befides to pro-

duce
will

it) if

none of
at

its

coefficients are

impomble,
the fame
it

have,

leaft,

one

real

root,"

with the root of the fimple equation whence

is

produced.

The

fquare of an

impomble quan-

tity

may be
x
,

a
tity

real, as the fquare of a* is but " the cube of an impomble quan-

is

dill

impoffibie," as

it

(till

involves the

fquare

root

of

negative

as,

V
d
3

2,

X
is

a'

X V

5* =
two

1,

plainly

imaginary.

From which
fimple

it

appears,

that

though

equations

involving

impomble
other,

expreffions,

multiplied by

one an-

expreffion

may give a product where no impomble may appear yet " if three fuch fim-,

ple equations be multiplied by each other, the

impomble

expreffion will not difappear in their

pro-

138
product."

A
And
its

Treatise
hence
it

of

Part II.

is

plain, that

though a
all

quadratic equation whofe coefficients are

real

rnay have

two

roots impoffible, yet

" a cubic

equation whofe coefficients are real cannot have

all its

three roots impoffible."

it

10. In general,
fible expreffions

appears that the impofin the

is

cannot difappear

equation
;

produced, but when their number

even

that

there are never in any equations, whofe coefficients are real quantities, fingle impoffible roots,

or

an odd number of impoffible roots,

that the roots

but
"

"

become

impoffible in pairs

and that " an equation of an odd number of

menfions has always one
11.
real root."

di-

" The

roots of equations are either^-

Jitive or negative

according as the roots of the

fimple equations w.hence they are produced are

pofuive or negative."

=.

h,

z=i

If

you fuppofe

%==..

c9

x ==

b,

d,

&c. then

fhall

x-{-a~o, x-\-bz=zo, x + c=iQ, x-{-d=:o, and

the equation x
will

+ axx + bxx + cxx + d

a,

== o

have

its

roots,

r,

d,

&C

negative.

But to know when the roots of equations

are pofitive

and when negative, and how

many

there are of each kind, fhall be explained in the

next chapter.

CHAP,

Chap.

2.

ALGEBRA.

139

CHAP.
Of
the

II.

Signs

and Coefficients of
S.

EQJJ AT ION
12.

T "% J

HEN

any number of fimple equa-

VV

tions are multiplied

higher!:

by each other,
equal to
and, the
is

it

is

obvious that the

dimenfion of the

unknown quantity in their product is the number of thofe fimple equations

term involving the hisheft dimenfion

called

the fir ft term of the equation generated by this

multiplication.

The term
unit,
is

involving the
lefs

next

dimenfion of the unknown quantity,

the greater!:

than

by
;

called the fecond

term

cf the equation

the term involving the next diis

menfion of the unknown quantity, which

lefs

than the greateft by two, the third term of the

equation,

&c.

And

that

term which involves

is

no dimenfion of the unknown quantity, but

fome known quantity,

the equation.

is

called the

lafi

term of

" The number cf

unit"

terms

is

the higheft dimenfion of the

ah: ays greater than unknown quantity by

is

And when
marked
the
in

any
its

term
place.

wanting,

an

cfterijk is
coefficients

The

Jigns

and

of equations will be underftood ,by

following

confidering

Table,

where the
fimple

140

Treatise

of

Partll.

iimple equations x
plied

a, x b, &c. are multiby one another, and produce fucceflively

the higher equations.

az=zo

X b=.o znx ax
*

bx + ob [zzO, X* =
'

a Quadratic,

xz

b\xx*-\-ac\xx abc=zO, ei 4-bcJ

X#

a-j

+ab^

Cub ic.

X* - i=o

+ abcd^ o, a Biquadratic.

+ &/I
\-cdi

X # -*

'*

abc
b I
c

-\-

abed
, ,

-\-

ac
%

abd

Ky>x*\-ad
\- ae -\-bc
-\~

d\

+ *&*/ abcde-=Q* f/"7

(
.

abe
acd.

^eJ

bd
be

ljVxx f

ade
ace

yx*

,-\-

beds

led
-f- ce

^r de

bed bee J
bde
cde

i3'

Chap.
13.
it
is

2.

ALGEBRA.

141

From

the inflection of thefe equations

plain, that the coefficients of the firft

term
the

is unit.

The
fum of

coefficient

of

the fecond
b,
c,

term

is

all the roots (a>

d,

e)

having their

figns changed.

The
of

coefficient

of the third term

that

is

the

fum

all the products

can be made by multir,

plying any
another.

two of the

roots (a, b,

d, e) by one

The
cf

coefficient of the fourth

term

is

the

fum
their
all

all the produces that

can be made by multiplying

into one another any three

of the

roots,

with

figns changed.

And
is

after the

fame manner

the other coefficients are

formed.
all the
by-

The
roots

lad term

always the produft of

having their figns changed, multiplied

one another.
14.

Although
it

in the

Table fuch iimple equa-

tions only are multiplied

by one another

as

have

pofitive roots,
efficients will

is

eafy to fee, that

"

the co-

be formed according to the fame

rule

negative roots."
-

when any of the fimple equations have And, in general, if x^px*

+ qx r = o reprefent any cubic equation, then

mall

p
;

be the fum of the roots

the products

q the fum of made by multiplying any two of

,

them

r the product of

all

the three
u,

and, if

P>

<l->

*,

+ h +
^?

&c. be the

coefficients

of the 2d, 3d, 4th, 5th, 6th, 7th,

fcfV.

142

Treatise

of

Part

II.

&c\ terms of any equation, then fhall p be the fum of all the roots, q the fum of the produces of any two, r the fum of the products of any three, s the fum of the products of any
four,
t

the

fum of

the products of any five,

fix,

u the fum of the products of any

15.

Gfc.
is

When

therefore any
it is is

equation

pro-

pofed to be refolved,

eafy to find the

fum
:)

of the

roots, (for

it

equal to the coefficient

its

of the fecond term having

or, to find the

fign

changed

fum of

the products that can be

made by multiplying any

of them.

determinate

number
the

But

it

is

alfo eafy

"

to find the

fum of
/>*

fquares, or of any powers, of the roots."

The fum

of the fquares

is

always

2j.

For calling the fum of the fquares B, fince the fum of the roots is p and " the fquare of the fum of any quantities is always equal to the fum
-,

of

their fquares

added to double the products

that can be

made by multiplying any two of

them," therefore p*

=B

-j- 2^,

and confequently

B
a*

==

p
z

2q.
z

For example,
y

+b+
is,
z

-v

c\
%

=
z

+ b + c + 2ab+2ac+2bc that p =B + 2g? And a + b + + df == a* + b + c + d -\-2Xab + ac + ad + k + bd+cd that again, p B + 2q of B == p 20. And fb
c
1
9

is

for any other

therefore,

number of quantities. In general " B the fum of the fquares of the

roots

Chap.
roots

2.

ALGEBRA.
the quantities

143

may
**

always be found by fubtradting iq

from p 1 known,
16.

p and

being always

fince they are the coefficients in the pro-

pofed equation.

" The fum of

is

the cubes of the roots of

any equation

equal to p*

Bp pq + 3r."
the excefs of the
tities

For B

%pq +
xp

3 ri or to

gives always

above the

fum of the cubes of any quantriple fum of the produces that

z

can be made by multiplying any three of them.

+ b*+ ab ac fax&+b+ c (=z q X p) = * + &* + c* $abc. Therefore B

Thus
a*
c
5

if the

fum of

the cubes
^r,

is

called C, then

iliall

~ qx p C
B
=.
1

(becaufe

After the

C~Bp qp + ^r p 2q) = p> 'jpq + 3 r fame manner, if D be the fum of

and
:

DpC qB+pr 4s
ps-T$t.

the 4th powers of the roots, you will find that

and

the 5th powers, then mall

EpD qC-\-rB
may be found
\

if

E be

the

fum of

And

after the

fame manner the fum

the

of any powers of the roots

the powers being obvious. 17.

progreffion of thefe expreflions of the

fum of

As

for the figns

it

of the terms of the

equation produced,
that the figns of
in the
all

appears from infpection

the terms in any equation

-|-

Table are alternately

and

thefe

equations are generated

nually x

a> x

by multiplying contib% x~c, x di &c. by one

another.


^Treatise
The
firft
is

144
another.

of

Part

IL

term

is
,

always fome pure

the

power of x y and

pofitive

fecond

The

a, power of x multiplied by the quantities And fince thefe are all negar, &c. , that term muft therefore be negative. tive,

is

third term has the products of any

(

thefe quantities
coefficient

a,

two of
its

<r,

&c.) for

caufe

which products
gives

are all pofitive, be-

+.

For

the

like
all

reafon,

the next coefficient, confifting of duels

the pro-

made by multiplying any

muft be negative
->

three of thefe

quantities,
fitive.

and the next po-

So that the

coefficients,' in this cafe, will

be pofitive and negative by x

turns.
<,"

But u
fince

in

this cafe the roots are all pofitive

= a?
af-

x =r

Cy

=d
."

e,

&c. are the

fumed fimple equations. It is " when all the roots are pofitive^
ternate^

plain then, that

the figns are al-

+ and

iS. But if the roots are

all

negative, then

x+axx-\~bxx+ exx + d,
prefs the equation to be

&C.

= o,
-

will exall

produced
,

whofe
all

terms will plainly, be pofitive

the roots of an

io that

" when
it

equation are negative^

is

plain

there will be no changes in the figns of the terms

of that equation" 19. In general,

tive roots in

there are as

many
to

pofi-

any equation

as there are

changes

in

the

figns

from

+
to

of the terms from

>

or
are
9*

arid

the remaining roots

negative.

Chap.

2.

ALGEBRA.
The Rule
is

145

negative."
fible roots

general, if the impof-

be allowed to be either pofitive or

negative.

20. In quadratic equations, the

are either
(x

two

roots

both pofitive, as in this

z=z)

aXx b

xz

ax bx

-\-

abz=. o,

where there are two changes of the figns

they are both negative, as in this

Or

(PTa X
where there

Wb =) +^ x + ah =
x*
is

o>

not any change of the figns.

Or

there

is

one pofitive and one negative, as in

where there
figns
laft
,

is

neceffarily
firft

one change of the

is

becaufe the

term

pofitive,

and the

negative, and there can be but one change

whether the fecond term be -f or Therefore the rule given in the 19th feclion
extends to
all

quadratic equations.
equations, the roots
this,

21. In cubic
i.

may
-{-

be,
c

All pofitive as in

x aXxbxx

rr o, in which the figns are alternately

-

and

as

appears from the Table

and there are

three changes of the figns.

2
.

The

roots

may be

all

negative, as in the

equation x

+ a x x -f b x x + c = o,
figns.

w here
T

there

can be no change of the

Or,

3*

146
3
.

-4

Treatise
as

of

Part
roots

IL
and

There may be two

pofitive

one negative,

in the equation x

axx-*b
o.

x-{-c

=o

which gives

-\-abl ~\- ab a1 z ac\ x b > x

-f-

abc

+
Here
there
if

c\

-bc\
the (Igns
:

muft be two changes of

a-\-b
is

becaufe

greater than

c,

the iecond

term muft be negative,

b-\-c.
ac

its coefTicient

being

a,

And if a -f- is lefs than r, then the third term muft be negative, its coefficient -\- ab

be (ab

+ b) *

being

in

that cafe

negative.

And

there cannot poffibly be three

fifft

changes of the figns, the

and

laft

terms

having the fame

fign.

4
x

There may be one

as in

pofitive root

and two

negative,

the equation x

=
c
ff
5

+a X

x-\-bx

o,

which gives
-\-ab)

-j-tf}

Where
figns, fince

-\-b

> x*

ac>x
term
is

abc

= o.
and the

e.y

be)
firft

there

muft be always one change of the

the
pofitive

laft negative.

And

there can be but one change

of the

figns, fince if the

fecond term
r,

is

ne-

gative, or a b lefs than -f* Becaufe the rectangle a

the third
lefs lefs

muft be

is

than the fquare

than a
-+-

& 4~ b

a -Y b> and therefore much

negative

Chap.

2.

ALGEBRA.
fo

147
be but one

negative alfo,

that

there will

change of the
is

figns.

Or,

if

the fecond
is,

term
there

affirmative,

whatever the third term

will

be but one change of the figns.

It

ap-

pears therefore, in general, that in cubic equations, there

are

as

many

affirmative roots as

there are changes of the figns of the terms of

the equation.

The fame way

of

reafoning

may
all

be

ex-

tended to equations of higher dimenfions, and!

the rule delivered in 19, extended to
equations.

kinds of

22. There are feveral confeclanes of

what
But

has been already demonftrated, that are of ufe

in

difcovering

the

roots

of equations.
it

before

we

proceed to that,

will

be convenient

to explain

fome transformations of equations, by which they may often be rendered more fimple, and the inveftigation of their roots-

more

eafy.

CHAP,

*4$

/Treatise

of

Part

IL

CHAR
Of
ons
;

III.

the Transformation of

Equatitheir in-

and exterminating

termediate terms.
23.

TT7E
Vv
:

now
and

proceed

to

explain

the

transformations of equations that

firft,

are moft ufeful

roots

"

The

affirmative

of an equation are

changed into

negative

roots of the

fame value ^ and

beginning

the negative roots

into affirmative,

by only changing the figns of the

terms

alternately,

with
x4

Thus

the roots of the equation

xz

the

fecond"
io,#
s

-J-49*

30 =
3

are

+ I? + 2, + 3, 5;

where-

as the roots of the

fame equation having only

the figns of the fecond and fourth terms changed,

mZt # 4 -j-#
2,.

ic)x

49* 30 = o,

are

--3,

+5.
this rule,
let

To
aflfume

underftand the reafon of an equation, as x

e,
.

us

dxx
-\-c,

axx bxx ex
+a
9

&c. ==
-\-e 9

o,

whofe roots are

+ j
&c.

+d,

&c. and another having

its

roots of the fame value, but affected with contra-

ry figns, as x

=0

+ axx + bxx + cXx + dxx + e^

is

It

plain,

that the terms taken alfirft,

ternately^ beginning

from the

are the

fame
ins

Chap.
in
*c

3.

ALGEBRA.
and have the fame
art

tig
fign^

both equations,

being products of

even number of the

roots j"

the product of any two roots having

the fame fign as their product

figns are

changed;

as

+ ax b~ ax+b.
all

when both

their

But the fecohd terms and

nately

taken alter-

from them, becaufe

will

their coefficients in-

volve always the products of an odd

number
in

of the roots,

have contrary figns

the

two equations.

For example, the product of

in

four, viz. abed, having the fame fign in both,

and one equation

the

fifth

abedx
that

>

and the other abed

term
e+
it

having
follows

their

product abede muft have contrary

figns in

the

two equations

thefe

two equaroots,

tions therefore that have the

fame

but

with contrary figns, have nothing different but

the figns of the alternate terms, beginning with

the fecond.

From which
is

it

follows,

"

that if

any equation

given^ and you change the figns

of the

alternate

terms,

beginning

with

the

new equation will have roots of the fame value* but with contrary figns."
fecondj the

24. It
equation

is

often very ufeful

tC

to

transform an
its

into
lefs

another

that fhall have

roots

greater or
it on

than the roots of the propofed eaua*

by fome given difference."

Let
#1

the

equation

-~p x z -{-qx
it

propofed

be

the

cubic

= o.
L

And
a

let it

be required to

transform

into another equation

whofe roots
mall

150
/hall

-4
be
lefs

Treatise
(e) 9

of

Part

II.

than the roots of this equation by

difference

ibme

and confequently x =jy e then y 9 inftead of # and its powers, fubftitute y e and its powers, and there will arife this new
,

xe

given

that

is,

fuppofe

equation,

(d)y*

+ $ey
=
Or

whofe roots are

lefs

than the roots of the pre(e).

ceding equation by the difference

If
it

had been required

to find

an equation

whofe roots fhould be greater than thofe of

the propofed equation by the quantity (e) 9 then

we muft have
quently x

=y

fuppofed y == x and confe5 e 9 and then the other equation

<?

would have had

this

form

(B)y>

3ey*

+ 3^ye

~ty-Y*py

pe>

If the propofed

equation be in this

form,

xz

+px

-\-

qx

r ==

o then by fuppofing x
4

+ =j
e
all

there will arife an equation agreeing in

refe-

-fbe&s with the equation

(A)

but that the

cond and fourth terms

will

have contrary

figns.

And

Chap.

3.

And

by

ALGEBRA. fuppofing x e

151

y, there will arife

all refpects,

an equation agreeing with (B) in

but that the fecond and fourth terms will have

contrary figns to what they have in (B).

The

firft

of

thefe

fuppoiitions

gives

this

equation,

(Qf 3tf\+ 3 y 2pey-\-pe

e
-jz

**
z

py

4- qy

~qe

o.

+r

The

fecond fuppofition gives the equation

-r

p/--i-2^y+KC
~f qy -r
qe

o.

25.
equations

The
is

firft

ufe of this transformation of

iC

to

ihew

bozv the fecond (or other

intermediate)
equation.'
9

term may be taken away out of an

It is plain that in

the equation
if

(A) whofe

fe-

cond term

is

%e

p x 7%

you fuppofe

e-=-\-p 9

and confequently 3^ term will vanlm.

In
the equation (C)

p = o,

then the fecond

whofe fecond term

e

is

#
J

3e

+ P X y\
the
z

fuppofing

= \p*

the fecond

term

alfo vaniflies.

px

Now

equation (A) was deduced from

-\-qx

rc

by fuppofing

y~x e:
and


'j

5$

A
-,

Treatise

of

Fairtll.
-\-px %

2nd the equation (C) was deduced from x 1 by fuppofing y x e. -f qx -f- r =z o

= +

From

which

this

Rule may

eafily

be deduced for ex-

terminating the fecond term out of any cubic

equacion.

RULE.
P Add
tion
to the

unknown

quantity of the given equacoefficient

the

third part of the

its

of the

fecond term with

proper JIgn, viz.

T rP*
'

and fuppofe this aggregate equal to a new unknown quantity (y). From this value of y find a value of x by tranfpofition^ and fubftitute this
value of x and
tion^
.

its

powers in the given equa-

and

there will arife a

new

equation that

Jhall

want

the fecond term/'

EXA
Let
'(term
it

M PLE.

be required to exterminate the fecond

out of this equation, x l

c>x

-\-

i6x
\

34

= o, fuppofe x

>

= y,

or

y+ =x
3

and

fubflituting according to the rule, you will find

9y 54y~Si(
2

3.4J

~~j

***

IO

=c

In

Chap.

3.

ALGEBRA.
is

153
is

In which there

no term where y
is

of two

dimenfions, and an afterifk

placed in the
it is

room

of

the fecond term, to fhew

wanting.

26. Let the equation propofed be of any number of dimenfions represented by (n) ; and
let the coefficient

of the fecond term with

its

fign

prefixed be

p, then fuppofing x

~-

=s y,

and confequently x
this

=y

-j-

an ^ fubftituting

value for x in the given equation, there

will arife a

new equation

that fhall

want the

fe-

cond term.
Chap.

from what was demonstrated in fum of the roots of the propofed equation is + p ; and fince we fuppofe
It
is

plain

2.

that the

= x
st

it

follows, that in the

lefs

new

equation,

each value of y will be

value of
roots
is

than the refpective

by
it

and, lince the

number of

the

#,

follows that the

of y will be

lefs

than

^>,

the

fum of fum of

the values
the values

of x y by n
that
is,

the difference of any two roots,

:

by

+p
is

therefore the
-j-

fum of

the va-

lues of

will

be

p p == o.
of the fecond term of the
the values of _y, viz.
coefficient
is

But the

coefficient

equation of y

+ p p,

the

fum of

and therefore that

;

equal

to nothing

and confequently,

in the equation

of

13'4
ofjy,

<A

Treatise
may

af
It

Part

IX,

the fecond term vaniihes.

follows then,

that the fecond term

.of

be exterminated out

any given equation by the following

RULE.
*'

Divide the

coefficient

of the fecond term of the

propofed equation hy the number of dimensions

of the equation
quantity y,

and affuming a new unknown

to
it

add

the quotient having its

Jign

changed,

equal to x the
pofed equation
fubftituie
,

^hen fuppofe this aggregate unknown quantity in the pro-

and for x and its powers* the aggregate and its powers^ fo
equation that arifes

Jhall the

new

want

its

fe-

cond term"

27. If the propofed equation

as

is

a quadratic,

px

-f q

rule,

fuppofe y

o, + p

then, according
i=z x->

to

the
this

and fubfrituting

value for x, you will find,

z

py ip 1\^
z
"

4 py 4- iP
4-

1
z

r
And
from
this

ip

+ q = p.
:

example the ufe of exterminatfor

ing the fecond term appears

commonly

the

fplution of the equation that wants the lecond

term

is

more

eafy.

And,

if

you can find the

value,

Chap.

3.

ALGEBR A,
new
%

155
it

value of y from this

to find the value of x

equation,

is

eafy

by means of the equation

%ririp *=
'

For example,
2

Since'

that

f %f and y V Ip* x y + Ip ip _ ^ if
q,
q_ ;

f -f a ~p

=z o,

it

follows

that
^,

fo

which agrees with what we demonftrated, Chapter 13.

Tart

L
%

If the propofed equation

x*

jfrr -jJVj

qx

is

a biquadratic, as

rx

-f-

s z=.

o, then

by fuppofing
the propofed

t?

or Xz=z y Jrip-> an equation fhall arile

having no fecond term.

is

And

if

of five dimenfions,

then you

mud

fuppofe

x=y+_ip. And 28. When the

tion
is

fo on.

fecond term in any equafollows, that

wanting,

it

"

the equation

has both
that the
to

affirmative

and negative roots," and

the

fum of the affirmative roots is equal fum of the negative roots by which
:

means
is

the coefficient of the fecond term,

all

which

the .fum of

the

roots

of both forts, va-

nifnes,

and makes the fecond term vanifh.

Ct

In

gen:ral\>
is

the

coefficient

of the fecond

term

the difference between the

fum of

the

affirmative roots
roots :"

and the fum of the negative

the roots

and the operations we have given ferve

ail

only to diminifh
the affirmative

when
is

the

fum of

is

greater!, or increafe the roots

when

the

fum of

the negative

greateft, fo

as

156
as

"A Treatise
.

c/

Part

II.

to

balance them,

and reduce them to an

equality.
It
is

obvious,

that in a quadratic equation

that wants the fecond term, there muft be one

root affirmative and one negative

and thefe

muft be equal to one another.

In a cubic equation that wants the fecond

term, there muft be either, two affirmative roots

equal, taken together, to a third root that muft

be negative
cc

or,

two negative equal

to a third

that muft be pofitive.

x qx r Let an equation x 3 o be fx propofed, and let it be now required to exter mi-

mi nate the third term." x By fuppofing y

third

=
2pe

e,

the coefficient of the

is

term

in the

equation of y

tion

A)

to be

3^

found

(fee

equa-

q.

Suppofe that co-

efficient

equal to nothing, and by refolving the

quadratic equation %e z
find

2pe
will

-f q

= o,

you
for

will
it

the value of

the equation

y=x

e,

which fubftituted
e,

in

fhew how to

trans-

form the propofed equation into one that fhall want the third term. The quadratic %e z o gives ipe -j- q

So that the propofed cubic

3
will

be transformed into an equation wanting the

third term

by fuppofing y

=x

?. .7~

.?f
>

*&?

+ = ** p" s/p*M \

If

Chap.

3.

ALGEBRA.
is

157

If the propofed equation

of n dimenfions,

the value of e 9 by which the third term

taken away,
.equation
ft

is
7.1)

may be had by refolving the quadratic

1Q
-\

+X e n ing p and + q to be
The

nxh i

= o,

fuppof-

the coefficients of the fe-

cond and third terms of the propofed equation.

fourth term of any equation

a cubic equation,

may

be taken
is

away by folving
coefficient

which

the

of the fourth term in the equation

as

when transformed,
this

in the fecond article

of

chapter.

The

fifth

term

may be taken

-.feme

away by folving a biquadratic ; and after the manner the other terms can be exterminated
if there are

any.
are other tranfmutations of equa-

29.
tions,

There

that

on fome occafions are

An

equation, as x z

px

ufeful.

qx

=
r

o,
its

may
roots

be transformed into another that Jh all have

equal to the roots of this equation multiplied by

given quantity , as /, by fuppofing

r

s= fx, and
this

confequently x
for

= -t
r

and fubftituting

value
arife

in the

propofed equation, there will

fi

71+ 7 =
.

o, and multiplying all

by

. .

fpy + pqy pr = o,
T

where the co-

efficient

of the fecond term of the propofed equa-

lion multiplied into

/ makes

the coefficient of

the

>

1.58

Treatise

of

Partll.
,

the fecond term of the transformed equation

and the following coefficients are produced by

the following coefficients of the propofed equation (as q 9 r, &c.) multiplied into the

powers of

f(f-,f,&c.)
Therefore " to transform any equation into
another whole roots fhall be equal to the roots
of the propofed equation multiplied by a given.
quantity
5 '

(/),

you

need

only

multiply

the
at

terms of the propofed equation, beginning

the fecond term, by /,

/%/

,/

4
,

&.c.

and put-

ting y inftead of x there will arife an equation having its roots equal to the roots of the pro-

pofed equation multiplied by (f)

30.

as required.

The
is
-

transformation mentioned in the

lad
the

article

of ufe when the higher! term of

has
it,

equation
;

coefficient

different

from
of

unity

for,

by

the equation
fhall

may

be tranf-

formed into one that

have the

coefficient

the highefl term unit.

If the equation propofed
=.
is

&x*

px
(a).

-\-qx

o,

then

transform

the

equation

into

one

whofe roots are equal to the roots of the propofed

equation
ax<>

multiplied
or x

by
,

That

is,

fuppofe y =1

and there
j

will arife

S -? + f ra =
z

*** &* + W
From

Oc

Chap.

3.

ALGEBRA.
eafily

159
this

From which we

draw

RULE.
<!

Change the unknown quantity x

prefix

into another y,

no

coefficient

to

the highefi term* pafs

the fecond* multiply the following terms* begin-

ning with the third* by a, or* a\

a\

dec. the

powers of the coefficient of the highefi term of ed the prop of equation , refpetfively."

Thus
is

yi yl

L%y
Then

the equation 3 a; 3 transformed into the

z

I3# 4-i4#'+.i6==0j
a

equation *

-{-

__ 13y-

+ 4 ov + I44 =
:

14

16
O.

9 2= o,

or

finding the roots of this equation,

it

will eafily be difcovered what are the roots of

the propofed equation

And
2,

therefore
it

fince

fince 3* =jy, or x one of the Values of

-fj.

is

follows that one of the values of x

is

31. By the lad Rule " an equation is

cleared

eafily

of fractions"
is

Suppofe
4l

propofed x 1

xz

x m
i)

x
11

or ~
c

the

equation
o.

Mul-

tiply all the

terms by the product of the deno-

minators, you find

*

*3

* nep x

#*'-f* *#

Ar =

o.

Then (by

laft feclion)

transforming the equation

into one that fhall have unit for the coefficient of

the higheft term,

nep

xy

+ ni^nq X v ffpffi&r = o.
Or,

you

find

i6o

A T r eat

xz
,

e of

Fart

IJL

Or, neglecting the denominator of the lad

term
tion

you need only -multiply

will

all

the equa-

by mn, which
z

give

mnx
-,

npx x + mq X x
,
.

mnr
?n
2

ss
n2r

o*

A And

.,

then y l

np X y + ffrq X y
after the values

= o.
it

Now

of y are found,
\

will

be eafy to difcover the values of x

firft cafe,

fince, in

the

*,

in the fecond,

=
reduced to

For example, the equation

x3 *
this

form jtf *
3

4#
4-tf
?

**

= i

is fir ft

27

= 0,

and then trans-

formed into j 146=0* 12V Sometimes, by thefe transformations, " Surds

arc taken away."

*
l

The

equation

x ps/a x x
y/a

As

for example,
z

-f J*

*V# = 6
is

by putting y
formed
--

?=:

x*

or # 3= -7-,

trans-

into this

equation,
q

p^a x +
l

-7-

r/*
make

== o.

"Which by multiplying

becomes y on free of furds.

ceed,

pay

+ q?y ra* o*
in order to

all

the

terms by ^-/^
an equatithis fuc-

But

the furd {j/a)

muft enter the alternate

3 2>

terms beginning with the fecond.

Chap.

3.

ALGEBRA.
as

i6x

32.
the

An equation,

xz

px + qx r=ro,
z

may he transformed
quantities
,

into one

wbofe roots jhall he

reciprocal

of

by fuppofing

and y

=
l

or, (by
z

one fuppofition)

x =z

becomes z

qz

-{-

prz

=
r
z
-,

o.

In the equation of j,

it is

is

manifeft that the

fo

order of the coefficients

inverted

that if

the fecond term had been wanting in the pro-

pofed equation,

the lad but one fhould have

been wanting in the equations of y and z. If the third had been wanting in the equation propofed, the
laft

but two had been wanting in the

equations of y and z.

Another ufe of
<6

this

transformation

is,

that

the great eft root in the one is transformed into

the leaft root in the other"

For

fince

=
y

and y
is

r it is

plain that

when
is

the value of

greateft, the

value

of/

leaft,

and con-

verfely.

How
all its

an equation

is

transformed fo as to have

roots affirmative, (hall be explained in the

following chapter,

CHAR

62
>

Treatise

rf

Part XL

CHAP.

IV.

Of finding the Roots of Equations when two or more of the roots

are equal to each other.
33.

TJEFORE we proceed to explain JD to refolve equations of

all

how forts, we
to

lhall firft

demonftrate " how an equation that

or more
roots
equal, is

has two

deprejfed

lower dimenfion 5" and

fequently,

its

refolution
fhall

made, con*

more

ealy.

And
of
this

endeavour to

explain
rules

the grounds
fhall

and many other

concife

we

give in the remaining part of this

Treatife, in a

more flmple and

manner

than has hitherto been done.

In order to
this,

we muft look back

if

to 24.

where we find that

-{-qx

any equation,

as

px

r=o,
it

is

propofed, and you are to tranfits

form
lefs

into another that mail have

roots

than the values of x by any given difference,

e,

as

you

are to alTume
its

ftituting for x

value

=x y + e you
*,
7

and fubfind th

transformed equation,

f 4- 3 y +
e
z
z

py

3 e *y 2P e y

+ t

e%

e%

o.

Where

Chap.

4.

ALGEBRA.
are

163

Where we
i.

to obferve,

That
x.

the laft term (e z

p^ +
qe

r)
e

is

the very equation that was propofed, having

in

place of
2*
is

3^

2pe +
by
1

The

coefficient

of the lad term but one

is

q}
z

which

the quantity that

laft

arifes

multiplying

coefficient e
in

pe
the'

every term of the

qe

-\-

r by the index of
e

each term, and dividing the product 3^

%

to

2pe
all

qe

by

quantity

that

is

common

the terms.
3.

is

y p

The
y

coefficient

of the lad term but two

the quantity that arifes

which

is

by
laft

multiplying every term of the coefficient

found

(3d'

2pe + q) by the index of

e in

each

term, and dividing the whole by 2e.

34. Thefe fame obfervations extend to equations of all dimenfions. If it is the biquadratic x*

pofed,

rx-\-szno then by iuppofing y = x

px
z

-\-

qx

that
e9
it

is

pro-

will

be

transformed into
j4_|. 4 gl

this

other,

py 3Po
l

J_ fay;Z

_|_

^y J^
2qey
ry

e 4r
z

"j

spfj pe

4-

qy

4 qe

> =1 o.
j

re
j

-f-s

Where

again

it is

obvious that the

laft

term
e

is

the equation that

was propofed, having

the laft

in

place of x.

That

term but one has

for

164
for
its

^Treatise
coefficient

of

PartIL
by

the quantity

that arifes

multiplying the terms of the

the indices

laft

quantity by

of

e
e.

in

each term,

and dividing
of the

the product by
laft

That the
z

coefficient

%pe term but two (viz. 6e q) is deduced in the fame manner from the term immediately following
,

that
z

is,

by multiplying every

term of 4^
of
e
e

is,

%pe

+
X

iqe

by the index
in

in

that term, and dividing the whole by

multiplied into the index of

the term

fought, that
is

p ==

by 6**X2

2.

lpex.1

And
.

the next term

The

demonftration of

this

may

eafily

be

made

general by the

Theorem

for finding the

powers

of a binomial,
confifts

fmce the transformed equation

are

marked by

of the powers of the binomial y e that the indices of e in the laft term, each by their coefficients
1,

multiplied

+ q,
fily

r,

+s

&c.

refpedtively.

35- From

the laft

two

articles

we can
term
is

ea-

find the terms of the transformed equation

without any involution.

The

laft

had

by fubftituting
equation
;

inftead of x in the propofed

the next term, by multiplying every

laft

part of that

term by the index of

e
-,

in

each part, and dividing the whole by

the following terms in the
the

and
in

manner defcribed
refpective

foregoing

article

the

divifcrs

being

Chap.

4.

ALGEBRA.
e

165

being the quantity

multiplied by the index of

in

each term.
demonftration for finding
will

The
this,

more. roots are equal

that

be eafy,

when two or if we add to

enters

" when

the

unknown

quantity

all the terms

lues
is

of any equation^ then one of its va-

equal to
z

nothing*
z=l

As

in

the equation

px

-{-

qx

o,

where x
follows

=
that

being

one x
%

px

of

the
z

fimple
it

equations

produce

+ qx = o,
is

that one of the

values of x
z

o.

In like manner two of the

values of x are equal to nothing in this equation

px = o
is

-,

equation x*
It

px

and three of them vaniih

%

in the

== o.
(converfely)

alfo

obvious
all

that

"

if

does not enter

i. e.

the terms of the equation,

if the laft

term be not wanting, then none

of the values of x can be equal to nothing

for if every

o cannot be a
x.

term be not multiplied by

diviibr

x, then

of the whole equation,

the terms

and confequently o cannot be one of the values

of
If x
z

does not enter into

all

of the equation, then two of the values of x

cannot be equal to nothing.
ter

If x l does not en-

into

all

the

terms of the equation,

then

three of the values of x cannot be equal to no-

thing,

&c.

36. Suppofe now that two values of x are equal to one another, and to e ; then it is plain
that

two values of y

in the

transformed equation
will

166
will

^
terms

Treatise
:

{/

Part II.

be equal to nothing

fince

e.

And
laft

confequently, by the lad article, the two

of the

transformed equation

mud

vanifh.

Suppofe
is

it

is

propofed, viz. x l

the cubic equation of 33. that px z -\-qx r =-. o ; and be-

caufe

we fuppofe x

e,

therefore the laft term of

e*

the transformed equation, viz.

will vanifh.

pe
z

z
-\-

qe

And
at

fince

two values of y vanifh,

viz.
r

the
will

laft

term but one,

But, by

;e

vanifh

the fame time.

2pey y So that %e z
e

x
\

qy

ipe

q 3= o.

fuppofition,
x,

=
"

therefore,

tion x 3 that

px = 3#* 2px
z

when two values of

-\-qx

in the equait

G, are equal,
o.

follows,

~j-

And

thus

the

propofed cubic

is

depreffed to a quadratic that has

one of

its

roots equal to one of the roots of that

cubic."

x4

px + qx rx + szd,
l

If

it is

the biquadratic that

l

is

propofed, viz.

and two of its roots

-'d=.

be equal
lues of
will

then fuppofing

#,

two of the va-

y muft

vanifh; and the equation of

this

34

be reduced to
y$
,

form,

J__ A_ey % -[- 6e-y z )

pyi

gpg L*i*
+
V' J
y
?'

So

that

4e

3P
~-

e%

4-Y

3fx*-\-

ro 2qx =
i-qe

or, faecaufe

e,

O.

In

Chap.

4-

ALGEBRA.
when two
e,

167

In general,

values of x are equal

the

to each other, and to

two lad terms of

:

the transformed equation

vanifh

and confe-

quently,

"

if

you multiply the terms of the

be

propofed equation by the indices of x in each

term, the quantity that will
arife will

= o,

and
fion
its

will

give an equation

of a lower dimen-

than the propofed, that fhall have one of

roots equal to one of the roots of the pro-

pofed equation."

niih

That the lafl two terms of the equation va when the values of x are fuppofed equal to
e>

each other, and to

fidering, that
fince

will

alio

appear by con-

two values of y then beis

come

equal to nothing, the product of the va-

lues of

y muft

vaniih, which
;

equal to the

lafl

term of the equation

alio that

and becaufe two of the

it

four values of y are equal to nothing,

follows

one of any three that can be taken out

of thefe four muft

products
vanifh
,

be-o;

and therefore, the

three

made by multiplying any

and confequently the
is

mult

coefficient

of the

lad term but one, which

thefe products,

equal to the

fum of

muft vanifh.
the

37.

After

fame manner,
and

if

there

are

three equal roots in the biquadratic x+

qx~-

px*

rx
;

-j- s

o,

if e

be equal to one of x

them

three values of
3

(=

e) will vaniih,

and confequently

jy

will enter all

the terms of
the

3.

i68
the
this

A
transformed

Treatise
equation
-,

of

Part
will

II.

which

have

form,

+^*** =
q == o
;

So

that here
therefore,

6^ ^
6#
2

^pe

p#

+
-\-

or, fince ?

= #,

and one of the roots of

this

quadratic will be equal to one of the roots

of the propofed biquadratic.

In this cafe, two of the roots of' the cubic
3 equation 4# t- 3p#

+ iqx r = o

are roots of

the propofed biquadratic, becaufe the quantity

6x

%px + deduced by multiplying 2qx

z

is

from 4# 3
the terms

^px

+
in-

r,

by the

dexes of x in each term.

In

general,

" whatever

is

the

number of

equal roots in the propofed equation, they will

all

remain but one

it,

in the equation that is deall

duced from

by multiplying
;

the terms by
all

and they will the indexes of x in them main but two in the equation deduced fame manner from That ," and fo of the

re*

in

the

reft.

38. What we obferved of the coefficients of e, equations transformed by fuppofing y x

leads to this eafy demcnftration of this

Rule

->

and

will

be applied in the next chapter to de*

the rules for finding the
limits

monftrate
equations.
It is

of

obvious however, that though

we make
the

yfe of equations whofe figns change alternately,

Chap.

4.

ALGEBRA.
all

169
other equa^

the fame reafoning extends to

tions.
It is

a confequence

alfo

of what has been

roots of any equaare equal, then
feries,

demonftrated, that
tion,
as x 3
z

"

if

px + qx r z= o,
+
ax 1
2#,

two

multiplying the terms by any arithmetical

as a

be

= o.
For
fince
z

%b, a

+
z

b 9 a y the product will

apx
bx
2

%x

2px
3^
3

+
o,
z

aqx
it

ar = o

and

## 3 -f
the

4- q

^/'^

2bpx + ### + bqx =

*?r

follows that
o.

Which

is

the product that ariies by multiplying

terms

of the propofed equation

feries,

by

the

terms of the

3^, a

+ 2b, a + b,

a\

which may reprefent any arithmetical progreffion.

CHAP.

170

Treatise

of

Part

II.

CHAP.
Of
39.

V.

the Limits of

Equations.
to

TT7E
VV

now proceed

(hew how to

diicover the limits of the roots of

is

equations, by which their iblution

cilitated.

much

fa-

Let any equation,

be propofed
the equation
;

as

x2

pe* -f qx r = o,
it,

and transform

as

above, into

f 4- -itf+'&y-k

pf 2peype
+
<LJ

?
.

+V
T
lefs

'

Where
fuppofe

the values of y are

than the refpece.

tive values
e

of x by the difference

If
all

you
the

to be taken fuch as to

make
%e

coefficients,

pe

of

the equation of y, pofitive, viz.

r,
%e"~

-\-qe

2pe-\-q,

p\

then
the

there

being
all

no variation of the
the values of v

figns

in

equation,

muft be negative;
f,

and confequently, the quantity

values of x are diminifhed,
the greateft pofitive value

by which the
be greater than
:

mud

of x

and confe-

quently muft be the limit of the roots of the

eauation x 3

px

-|-

qx

r == o.
It

Chap.

5-

ALGEBRA.
Ci
l

171

It is fufficient therefore, in

order to find the

limit,

to

enquire what quantity fubftituted for

in
z

3*

2px +

each of thefe expreflions x

q,

%x

px

+ qx r,
all

pi will give them

po-

fitive

," for that quantity will be the limit re-

quired.

How
laft

thefe expreflions are

formed from one

another, was explained in the beginning of the

chapter.

EXAMPLE.
40.
If the equation x*
1

2x* iav
-,

+ 30tf"

+ 63X +
number
tion,
lt

20

=
-,

is

propofed

and

it is

required

to determine the limit that

is

greater than any

of the roots

you

are to enquire

what integer

fubftituted for x in the propofed equa-

and following equations deduced

from

y 35>

w^

8* ve >

eacn 5 a pofitive quan-

tity.

5# 4
5*
3

5**

ox* + 6ox + 63 -6** i$x + 15 \x 5

8x>
3 2.

5* ~~

The
limit

leaft

integer
is

number which
2
;

gives each
is

of thefe pofitive,

which therefore

the

of the roots of the propofed equation;

or a number that exceeds the greateft pofitive

root.

If

172

Treatise

of
is

Partll.
required,

If the limit of the negative roots

you may, by
their limits.

23,

change the negative into

in

pofitive roots, and thtrn proceed as before to find

find that

S
are

that the five roots of the propofed equation

3

betwixt
3

Thus,
is

the example,

you

will

the limit of the negative roots.

and

-J- 2

>

41. Having found the limit that furpafTes the greateft pofitive root, call it m. And if you

afTumej ==
pofitive

m x,

and

for

fubftitute

m y

the equation that will

j

arife will
is

have

all its

roots
all

becaufe
t

fappofed to furpafs

the values of x

and confequently

x (= y)

mufl always be affirmative. And by this means, any equation may be changed into one that jhall
have
all its roots affirmative.
if

Or

n reprefent the limit

fhall
its

of the negative

roots, then

by arTuming^:=;# + #, the propofed

be transformed into one that
,

equation

fhall

have

all

roots affirmative

for

+
it

being
follows,

greater than any negative value of x,

that,

x+n
" The

mufl be always

pofitive.

42,

greateft negative coefficient of any

equation increafed by unit, always exceeds the greateft

root of the equation."

qx r = o be propofed;
firft.

To

demonflrate

this,

let

the cubic x*

fx

where

all

the terms

are negative except the

it

Affuming^= x

will be

transformed into the following equa-

tion.

Chap.

5.

ALGEBRA.

qy

173

'

i.

are equal to each other

e =:

Let us fuppofe that the coefficients p, q, r, , and if you alfo fuppofe

i,

then the

laft

equation becomes

+ af
"Where
all

+3py

+ 3y
are
all
1
j

1=
it

the terms being pofitive,

follows

that the values of

negative, and that

is

confequently

e,

or

greater than

the

greater! value of
2.

x in the propofed equation.

lefs

If q
e

and r be not =j), but

fcili

than

it,

and for

you
(

fubftitute^

(fince the nelefs,

gative part
pofitive

becomes

the
all

remaining undiminifned) a

fortiori,

the coefficients of the equation (A)

fitive.

become poa

And

the fame

is

obvious

if

and r have
as
if,

pofitive

figns,
It

and not negative

appears therefore,

figns,

we
in

fuppofed.

"

that,

any cubic equation, p be the

coefficient,

greateft negative

then

+ mud
1

furpais the greateft

value of

x"
3*-

43-

By

the fame reafoning

it

appears,

that if q be the greateft negative coefficient of

the

174

<d

Treatise
e

of

Part

II.

the equation, and

= q -f
in

i,

then there will be

for

no variation of the figns it appears from the

the equation of
that

laft article,

if all

the three (p 9 q, r) were equal to

one another,

and
unit,

equal to any one of them increafed by

as

to q

i,

then

all

the

terms of the

equation (A) would be pofitive.

Now

if e

be

fuppofed
leis

ftill

than

q,

equal to q i, and p and r to be then, a fortiori, all thefe terms will

be

pofitive, the negative part,

which involves

and r being diminiflied, while the pofitive part

and the negative involving q remain as before. 4. After the fame manner it is demonftrated,
that if r
is

the greateft negative coefficient in the

e is

equation, and

fuppofed

=
-f- 1

-f-

i,

then

all

the terms of the equation (A) of

sitive
;

will

be po-

and confequentiy r
x.

will

be greater

than any of the values of

xl

px

What we
z

+ qx r =: o,

have

faid

of the
is

cubic

equation

eaiily

applicable to

others.

In general,

we conclude

that

"

the greateft

negative coefficient in any equation increafed by

unit,
is

always a limit that exceeds

all

the roots

of that equation."

But
the
unit,

it is

to be obferved at the

fame time, that

increafed
:

greateft
is

negative

coefficient

by

very feldom the neareft limit

that is
article.

beft difcovered

by

the

Rule

in the

39th

44*

Chap.
44.

5.

ALGEBRA.
in 41.
.

175
to

Having fhewn
;

how

change
have

any propofed equation

all its

into one that fhall

roots affirmative
all their

we

fhall

only treat of

fuch as have

roots pofitive, in

what

re-

mains relating to the limits of equations.

Any

fuch equation

may

be reprefented

by

x axx bxx cxx

roots are a
by
all
c,

d,

&c.

o,

whofe

d,

&c.
equations

And
eafiiy

of

fuch

two

limits

are

difcovered
is

from what precedes,

leaft,

viz. o,

which

lefs

than the

and

e^

found ac-

cording to 39. which iurpafies the greatell root of the equation.

But
tc

befides

thefe,

we

fhall

now

iliew

how

to find other limits

betwixt the roots themf elves"

And,
on
it

for this purpofe, will fuppofe a to be the

leaft root,
,

b the fecond root, c the third,

and

(o

being arbitrary.
If

45.

you

fubftitute

in place

of the un^

known

quantity,

putting x

o,

the quantity
is

that will arife

from that fuppofition

all

the laft

term of the equation,

the others, that involve

x 9 vanifhing. If you fubftitute

leaft

for X a quantity lefs than the

root a, the quantity

laft

refuking will
;

have
be
is

the fame fign as the

term

that

is,

will

pofitive or negative according as the equation

of an even or odd number of dimenfions.

all

the factors x

a,

by

For
will

c,

&c.

be negative, and their product will be

pofitive

or

176

^Treatise

of

Partll.
is

or negative according as their

number

even

or odd.
If

you

fubftitute for

x a quantity greater than

lefs

the lead root a,

roots,

but

than

all

the

other

then the fign of the quantity refulting

will be contrary to

a) becomes now pofitive, one factor (x others remaining negative as before. the
If

what

it

was before

becaufe
all

you

fubftitute for
leaft

x a quantity greater than

x

the two

roots, but lefs than all the reft,

both the factors x

a,

and the

reft

remain
will

as they were.

become So

pofitive,

that the

whole product
laft

have the fame fign as the

term of the equation.

Thus

fucceffively

placing inftead of x quantities that are limits

betwixt the roots of the equation, the quanti-

and
ties

ties that refult will

.

have alternately the figns

-\-

And,

converfely>

"

if

you
x

find quantiin the pro-

which

fubftituted in place of

pofed equation, do give alternately pofitive and

negative remits, thofe quantities are the limits

of that equation."
It
is

ufeful

to

obferve,

that,

in

general,
for

"
x

when, by fubftituting any two numbers

in

any equation,

the

refults

have contrary

figns,

one or more of the roots of the equation

muft be betwixt thofe numbers."

the equation x l

2x 5 =
z

Thus,

in

0,

if

tute 2 and 3 for x, the refults are

you
5,

fubfti-j-

whence

it

follows that the roots are betwixt 2

and


AL
for

; ;

Chap.
and 3
:

5.

GE BRA.
thefe
refults

177
have different

when

figns, one or other of the factors which produce the equations muft have changed its fign

fuppofe

it is

e,

then

it is

plain that

muft be

betwixt the numbers fuppofed equal to

z px z -f- qx 46. Let the cubic equation x o be propofed, and let it be transformed, r

x.

by affuming y == x

e,

into the equation

pf

2P e y

P r
e^

e%

o.

Let us fuppofe
and becaufe the
vanifh in
will
all

equal fucceiTively to

the

three values of x9 beginning with the leaft value

laft

term

pe +
%

qe

r will

thefe

fuppofitions,

the equation

have

this

form,
a

+ + f 3<? 3'
py

2pe >
1

=o
q

+
where the
laft

>
is,

term

3<?

ipe +
is

from the

nature of equations, produced of the remaining

values of y, or of the exceffes of two other values of x

above what

fince always
i.

y=x

fuppofed equal to e\

e.

Now,
they

If e be equal to the leaft value of #, then

thofe
will
3**

two

exceffes

being both pofitive,

give a pofitive product, and confequently

2pe+q

will be, in

this

cafe, pofitive.


178
2
.

Treatise
excefTes

of

Part II.

If e be equal to the

fecond value of x,

then, of thofe

two

one being negative

and one
will

pofitive, their

product

3^

2pe

q,

be negative,
If e

3.

be equal to the third and greateft

value of x, then the two excefTes being both

negative, their product

3^ 3^
e
3

ipe +
-f-

is

pofitive.

Whence,
If in

the equation

fubftitute fuccefiively in

roots of the equation

ipe = you the place of the =

q o,
e,

three
o,

pi" -f qs

the quantities refulting will fuccefiively have the

figns

+,

+
is,

and confequently the three

roots of the cubic equation are the limits of the

roots of the equation

^r

2pe -f q == o
is

(by

45.)

That
is lefs
;

the leaft of the roots of the

cubic
other

than the lead of the roots of the

a

the fecond root of the cubic

limit

between the two roots of the other;

greateft root

and the

of the cubic

is

the limit that ex-

ceeds both the roots of the other.

47.

We

have demonftrated that the roots

e

of the cubic equation

limits of the quadratic

pe 3^
3 3

-J-

qs

=o
r
\

are
it

ipe

+q

whence

Follows
dratic

(converfely) that the roots

of the qua-

3^

the

firft

2pe o are the limits between q and fecond, and between the fecond and

+ =

third roots of the cubic e

pe*+
qe

r == o.

So

that if

you

find the limit that exceeds the

greateft

Chap.

5.

ALGEBRA,

179

of the cubic, by 39. you will have (with o, which is the limit lefs than any
greater! root

of the roots) four limits for the three roots of

the propofed cubic.
It

was demon ftrated

l

dratic $e

2pe
e>

by multiplying each term by the index of e in it, and then dividing the whole by e , and what we have
pofed cubic
=-

pe

+q
z

is

+ qe r

in 3$. how the quadeduced from the pro-

o, viz,

demonftrated of cubic equations

is

eafily

extendthat the

ed to
laft
is

all

others

fo that

we conclude, "

term but one of the transformed equation

determining the limits of
Or, that the equation

the equation for

the propofed equation."

arifing

by multiplying each term by the index

in
it,

of the unknown quantity

is

the equation

whofe roots give the

equation
,

limits

of

the propofed

if

you add to them the two men-

tioned in 44.
48. the root
(i. e.
-Z-p)

For the fame

reafon,

it

is

plain thaC

of the fimple equation $e

is

pz=zp

the quadratic 3**

J

2pe And, = of 4* + pe + equation %pe == o quadratic two of the cubic mits betwixt 2pe + ~ o and 4^ p = o

the limit between the

-f q

two

roots of
as

o.

3pe

2( e l

gives three limits

re -f" s == >

the
the

e*

1^"

6e

-f a

gives

li-

that

are

the roots
,

4<?

iqe

gives one limit that

is

betwixt the two roots of

the

180
the

^Treatise of quadratic 6e $pe + q = o.

z

PartIL
So
that

we

have a compleat

feries

of thefe equations arifing

to the propofed, each of
limits

from a fimple equation

which determines the
equation.

of the following

49. If two roots in the propofed equation are equal, then " the limit that ought to be be-

twixt them muft, in this cafe> become equal to

perfectly agrees with

one of the equal roots themfelves." Which what was demonftrated in

lafb

the

chapter, concerning the

Rule

for finding

the equal roots of equations.

And,

the fame equation that

gives the

li-

mits, giving alio one of the equal roots,

two or more are equal, it appears, that you fubflitute a limit in place of the unknown
quantity in an equation, and, inftead of a pofitive

when " if

or negative refult,

it

be found == o, then

you may conclude,

-felf is a

that not only the limit it-

root of the equation,

but that there

it

are

two

roots in that equation equal to

and

to one another."

50. It having been demonftrated that the 5 roots of the equation o r qx px z

;>;

to

=
r,

%x z ipx +
-

are

the

limits

of the roots

of the equation

q = o,

the three roots of the cubic

equation, which fuppofe

be

tuted for x in the quadratic $x*

ipx +

a,

fubftiq,

muft
alter-

give the

reiults

pofitive

and negative

nately.


Chap.
nately.

5.

ALGEBRA.
;

jSi

My + L p ipa + = 3^ 2pb + q~ M 3^ 2pc + q=zL\ and pa*-\-qa = and 2pa =zNx<?<, fubtra&ing the former multiplied remainder to from pa 2qa In fame manner ^ 2^ + ^r = JVx Mxb, &ndpc 2qc+2r = +Lxc + fuch quantity Therefore px 2qx + %r
-

Suppofe

thefe three refults to be -f -A7*

is,
z

that

jV,

fince

<2.

o,

3<2

-\-qa in-

the latter, the

0.

is

the

3/"=:

7,

is

that

if,

for x>

you
a,

fubftitute in

it

fucceffively
b,

the refults will be

+ Nx a,
are

MX

a9

b, c y

+ Lxc
are

Whence
px*

b,

c,

limits

of the equation
converfely, the

2qx-\-^r = o (by 45.) and,

between the
firft

roots of the equation px*

limits

2qx + y = o

and fecond, and beequation /w*

tween the fecond and third roots of the cubic

px*+ qx r=zo.
=o
arifes

Now the

2qx-{-%r

from the propofed cubic by

multiplying the terms of this latter by the arithmetical progreflion o,

in the

fame manner

roots of the

are limits

And 4^ = equation p* 2qx px rx of the equation

1,

2,

3.

it

may

be fhewn that the

z

-\-%rx
1

x*

-f-

qx

+ = o.
s

Or,

multiply the

terms of the equation

X
by a

px

-}-

qx

r=o
-ftf,

<2# ?

+ 3^, a + z, a , apx + ar (= o) + 3^# 2^# + ^*'(=3* ~- 2px + qx&x).

%

^
a

'

Any


182

A
you

Treatise
difference,

of
is

Part

II.

Any
(if

arithmetical feries where a

the lead term

and b the

common

and the products

fubflitute

for

#,

fucceffively, a, b, c,

the three roots of the propofed cubic) fhall be

-{-NxbXi

Mxbx,

-\-Lxbx.

For the

firft

part of the products

a, b,
r,

x**

px

+qx

being limits

in the

equation %x l

2px-\-q
M,

= o;

and

= o,
L,
n

their fubftitution

muft give

refults iV,

alternately pofitive and negative.

In general^

px

+ qx'"

the roots of the equation x n

rx

+
3

&c.
n

= o are limits of
*

the roots of the equation nx

Xpxn ~ z

--n2 X

qx n i

X
is

rx"-4-\-, &c>

o
it

or of any equation that

deduced from

by

multiplying
gression a

its

terms by any arithmetical pro-

+ b,

a + 2b, a + 3b, a + 4.b, &c.

And

converfely^ the roots

of

this

new

equation will

be limits of the propofed equation

xn
<c

px" 1

-{-

%Xn ~*

&C.

= O.
For
as
(in

If any roots of the equation of the limits

then

are impojjible^
the

muft there be fome roots of

impcffible"
z

propofed

equation
%e

46.) the quantity

fcrated to be equal
cefTes

2pe +

was demon-

to the product of the ex-

of two values of x above the third fupe

j

pofed equal to

if

any impoflible exprefiion be

con-

found

in thofe exceffes, then there will of

fequence be found impoffible expreflions in thefe

two values of av

And

Chap.

5.

ALGEBRA.
this

183

And " from

obfervation rules

may be

deduced for difcovering when there are impofOf which we fhall fibie roots in equations."
treat afterwards.

51. Befides the there are others

method already explained,

limits

by which

may

be deter-

mined, which the root of an equation cannot

exceed.

Since the fquares of

affirmative,
it

all

real

quantities
the

are
the

follows,
roots

that

"

fum of
mufi

fquares
greater

of the than

of

any

equation

be

the fquare

of

the great eft

root."

And

the fquare root of that

fum

will therefore

be a limit that muft exceed the greateft root

of the equation.
If the equation propofed
is

xn

&c. == o, then the fum of the fquares of the roots (by 15.) will be p z iq. So that
n
3

rx
*Jp
z

px ~
n

-\-qx n

-j-,

iq
if

will

exceed the greateft root of that

equation.

Or

you

find,

by 16. the fum of the 4th

root of that fum,
it

powers of the roots of the equation, and extract the biquadratic

alfo
will

exceed the greateft root of the equation.

52. If you find a mean proportional between the fum of the fquares of any two roots,
a, b,
this
z

and the fum of

their biquadrates

(V + 4 ),
:

mean

proportional will be
b
6
.

^a^a'^+a+b' +

And

the

fum of the cuba

184
cubes
is

A
if

Treatise
Now
b,
it it

of
d"

Part IL
l

the fquare of a

a*+bK

fince

2ab-{-b
^
;

is
->

muft be always pofitive by a*b

alfo
%

and
a*b
z

you multiply

the product

zatP + a*b
z

4 will

be pofitive

and

confequently

a^+tfb*
Add
a
6

will
6
y

be always greater
G

than ia b\

+b

and we have a
6

+ a^
6
;

. a b

4-

^
3

greater

than a 4- 2a* b*
2.
'

+b

and extracting the root

greater than # 4- b
z
.

V ab _f #v III ^+ IJ* ^ And the fame may be

number of
roots

demonftrated
ever.

of any

what-

Now

if

you add the fum of

give

all

the cubes

taken affirmatively to their fum with their proper

figns, they
will

double the

fum of

the

cubes of the affirmative roots.

fubtracl the fecond
will

And
the

if

you
there

fum from
it

firft,

remain double the fum of the cubes of the

negative roots.
the

Whence
mean

follows, that

"

half

fum of fum of the

drates,

the

proportional betwixt the

fquares and the

fum of

the biqua-

and of the fum of the cubes of the

roots with their proper figns, exceeds the

fum
and
the

of the cubes of the affirmative roots :"

"

half their difference exceeds the

fum of

cubes of the

negative roots.

'

And by

ex-

trading the cube root of that fum and

rence,

diffe-

you will obtain limits that (hall exceed the fums of the affirmative and of the negative
roots.

And

fince

it

is

eafy 3

from what has

beea

Chap.

5-

ALGEBRA,
it

185

been already explained, to diminifh the roots

of an equation fo that they

gative but one,

appears

all may become nehow by this means

you may approximate very near to that root. But this does not ferve when there are impoflible roots.

Several other Rules like thefe might be given

for limiting the roots of equations.

We
z

fhall

give one not mentioned by other Authors.

In a cubic x z

px*+qx r=o
then
fhall the

find q

-J7,

2pr,

and
the

call

it

e^;

greateft root of

equation always

be greater than -j-, or

J
-J-,

And,
3

In any equation x"

px*
2pr+2s

*-{--

9&~*

r# ~
w

&c. -= o

find

an ^ extracting the

root of the fourth power out of that quantity,

it

fhall

always be

lefs

than the greateft root of

the equation.

N4

CHAP.

86

^Treatise

p/

Part II

CHAP.
Of

VI.
all

the Refolution of Equations,

whofe Roots are commenfurate.

was demonftrated, in Chap. 2. that JL the laft term of any equation is the product of its roots from which it follows,
:

5j.TT

that the roots of an equation,

rable quantities,
will

when commenfu-

be found

among

the di-

vifbrs of the laft term.

And

hence we have for

the refolution of equations this

RULE.
Bring
all the terms
to

one Jide

of the equation,
term, and fub-

find all the divifors of the

fiitute them
tity

laft

fuccejfively for the

unknown quan-

in

the

equation.

So jhall that divifor

which, fubfiituted in this manner* gives the

refult
tion.

=1 o, be the root of the propofed equa-

For example, fuppofe.

refolved,

this

equation

is

to

be

$ax bx

z
z

r{-j-

2a z x

2a

b !___

%abx
is

1
lite-

where the lad term

%a x b, whofe fimplc

ral

Chap.

6.

ALGEBRA.
:

187

ral divifors are a^ b, 2a, 2^, each of which may be taken either pofitively or negatively but as here we find there are variations of figns in the

equation,

we need only

take

them

pofitively.

Suppofe
ftituting

9C

divifors, and fuba for #, the equation becomes

firft

a the

of the

u.T
So
is

C or >

3*-3*+3**-3*
it

= o.

that, the

whole vanifhing,

follows that a

one of the roots of the equation.

After the fame manner,
if
is

you

fubftitute b in

place of x, the equation

b%

^ab

2a z b

ia b 7
z

___

which vanifhing fhews b to be another root of

the equation.

Again,

if

you

fubftitute

2a for

#,

you
be

will

find all the terms deftroy one another fo as to

make

the

fum

= o.
\2tf
%

For
?

it

will then

Za>

+ 4# ia*b 7 __ \a b + 6a b 5
~
z

Whence we
the equation.

find

that

2a

is

the third root of

Which,
the

after the firft

two

(.+ a,
col-

+ b)
lected

had been found, might have been

from
this, that
laft

term being the probeing known,

the

duel of the three roots,

+ a, + b

i88

^Treatise

of

Partll.

the third muft neceftarily be equal to the lad

term divided by the .product ab> that

ia b ab
z

is,

Let
x3

ix 33* + 90 = o
z

the roots of the cubic equation

be required.
are

And
J
}

firft

the divifors of

90

found to be

2, 3, 5, 6, 9, 10, 15, 18, 30,

1

45, 90.

fubftitute

for x,

you
1

will find x*
is

4-90
tion.

= 56;
:

2# 33*
a

If

you

fo that

not a root of the equa.

If you

fubftitute 2 for #, the refult will

be 24

^ So

-2,v

but, putting x 3, you have ~33>r4-90=:27-i8-99+90=:ii7-ii7=o.is

that 3

one of the roots of the propofed

other affirmative root
it,

equation.

The
you

and

after

find

as

it is

5 ; manifeil from the

is

is

-[-

equation, that the other root

negative,

you

are not to try any more divifors taken pofitively,

but to fubftitute them, negatively taken,

for x
loot.

and thus you

find that

For putting x

= 6,

is

the third

you have

-2 1 6- 72+ 198 90 ^ 3 ~2#*~33tf-f 90 This laft root might have been found by
viding the
laft

= 0.
di-

term 90, having its fign changed, by 15, the product of the two roots already
found.

5$.
is

When

found, in order to find the

one of the roots of an equation reft with lefs trou-

ble, divide the

propofed equation by the fimple

already

equation which you are to deduce from the root

Chap.

6.

ALGEBRA,
will

189

already found, and the quotient mall give an

equation of a degree lower than the propofed

whofe roots
quired.

give the remaining roots re-

As

gave #

3 = x ix x
=3
3)

for

example, the root

or #
z

3,

firft

found,

0,

whence dividing thus,

33-v-f-

90

{x -\-x

30

3.V

xz

33 x +9

3# %ox + 90 30*+ 90
o

The

quotient mall

give a quadratic equation

x z -\-x
cubic

which muft be the product of the other two fimple equations from which the
0,
is

30 =

generated,

and whofe roots therefore

of that quadratic equation
I.

muft be two of the roots of that cubic.

Now
and

the roots

are eafily found

6.

by Chap. 13. Part

to be

-L5

For,

x zj

add i 4
si y

= 3> A- ^-J--=3oJ-- = ^4 a 4
r *
'

+l

and ....#

+ =

4*

*,

or

6. 1

190

A
;

Treatise

of

PartIL

56. After the fame manner, if the biqua25^*-jr- 26^-^-120 2x 2 dratic x A is to be

refolved
for #,
vifors,

by fubftituting the
will find that

divifors

of 120

you
is

+
;

3,

one of thofe di"of

54 225-}- 78 + 20 = dividing the pro79 279=0. And you x pofed equation enquire 22X 40 = of the cubic x + x the
3 for x giving 81
2
1

one of the roots

the fubftitution

therefore
3,
3

-by

niuft

for
o,

7,

roots

and finding that

is

-~- 5,

one of the divifors of 40,

,v 5, and the quotient gives the quadratic #*_j_6#-L8 o, whofetwo roots are 2, 4.

one of the roots, you divide that cubic by

So

that the four roots of the biquadratic

3> -1-5*

are

2,-4.
-,

$y. This Rule fuppofes that you can find which you may all the divifors of the laft term

always do thus.

" If

it

is

a Jimple quantity , divide

r

it

by

its

leaft

divifor

that
its

exceeds
leaft

unit^

and

the

quotient

again by

divifor^ proceeding thus till

you have a

quotient that is not divifible by any

number greater

than

unit.

This quotient,
firfl

with thefe divifors, are the

vifors of the quantity.

or Jimple di-

And
2,

the

products

of the multiplication of any

3, 4,

&c. of

them

are the compound divifors"

;

As, to find the divifors of 60

firfl

divide

by

2,

and the quotient 30 again by

2, then the

next

Chap.

6.

ALGEBRA.
is
,

9r

next quotient 15 by 3, and the quotient of this

divifion 5

not farther divifible by any integer

fo that 2, 2,
.

above units

The The The The

The

fimple divifors are

3,

5.

products of two,

4, 6,

10,

15.

products of three,

....

12, 20, 30.

.
.

product of

all

four,

60.

divifors

of 90 are found manner.

after

the fame

Simple

divifors,

2,

3,

3,

5.

The The The

products of two,
products of

.... 6, 9, three, .... 18,'

four,

10,

15.

30, 45. 90.

product of

all

The

divifors

of 21 abb.
3,
7, a, b, b.

The fimple divifors, The products of two, 21, The products of three,

3^, 3^, 70, jb, #,

M.

210, 2i, $ab 9 %bb 9

yaby jbb, abb.

The The

products of four,

nab

2ibb, %abb, yabb. 2iab.

product of the

five,

58.

But

as the

lad term

may have

very

many

divifors,

and tne labour may be very

great to
quantity,

fubftitute

them

all

for the

we

fhall

now fhew how

abridged,
divifors

by limiting to a fmall

unknown may be number the

it

you are to try*. And firft it is plain, from 42. that " any divifor that exceeds the
greater!

t92
greatefl:

-^Treatise
Thus
z

of

PartIL
is

negative coefficient by unity

in

to

be
the

neglected."

x*

2# 2$x
3

refolving

the equation
as

-{-26x

+ 120 = 0,

25

is

greatefl negative coefficient,

the divifors of 120 that

glected.

we conclude that exceed 26 may be ne-

But the labour may be (till abridged, if we make ufe of the Rule in 39 that is, if we find the number, which fubitituted in thefe fol-,

lowing expreffions,

3# 25* 6x 6x 2X
2* 5
a
%

tf

2tf 3

2$X
25,

-\~j-

l6x
13,

-J-

120,

I,

will give in

them

all

a pofitive refult

for that

number
and
all

will

be greater than the

greatefl: root.,
it

the divifors of 120 that exceed

may
we
y

be neglected.

That

this

invefligation

may be
with
that

eafier,

ought to

begin

always

expreffion,

where the negative roots feem

where finding that 6 fubftituted
expreffion pofitive, and gives
preffions at the

to prevail

as here in the quadratic expreffion

6x z
for

6x 25;
x gives that
conclude

moft

all

the other ex-

fame time

pofitive, I

that 6
all

is

greater than any of the roots, and that

the divifors of

120 that exceed 6 may be

If

neglected.

Chap.

6.

If the

A L G E BRA. 193 equation x + ix + iox 72 = o is

z

propofed, the Rule of 42 does not help to abridge the operation ; the laft term itfelf be-

ing the

greater!:

negative term.

But, by 39.

we

enquire what
all

number

fubftituted for
:

will

give

thefe expreflions pofitive

xz
Z

-J- 1

7 ix -(- io,v
-

72,
we need
fee

%X -{-22**
3X

~j-

10,

+.11is

Where

the labour

very fhort, fince

expreflion
-,

only attend to the

firft

and we

immediately that 4 fubftituted for x gives a politive refult, whence all the divifors of 72 that
exceed 4 are to be rejected
trials
;

and thus by a few

the pofitive root of

we 'find
2,

that

+2

is

the equation.

x
is

Then

dividing the equation by

and refolving the quadratic equation that

the quotient of the divifion,

two roots to be
59.

9,
is

and

you
4.

find the other

But there
to

another method that relaft

duces the divifors of the

ufeful,
(till

term, that can be

more narrow

limits.

Suppofe the cubic equation x*

px

=o
And
will

+ qx r
it

is

propofed to be refolved.

Transform
*lefs

to an equation whofe roots fhall be

the values of x by unity, afiuming v

the

=xu
jy,

than

be

p-{-q

laft

term of the transformed equation

r
-,

which

is

found by fubfor

ftituting unit, the difference

of x and

x9
in


194
in

-^

Treatise
,

of

PartIL
appear

the propofed equation

from 24. where,

when y=zx

as will eafily
e,

of the transformed equation was

Transform again the

-

pe*+qe px*+qx equation

e*

the laft term

r.

a?

= o,

by afluming y

=x+

1,

into an equa-

whofe roots mail exceed the values of x by unit, and the laft term of the transformed 1 equation will be r, the fame p q
tion

1,

that arifes

by

fubftituting

the difference

betwixt x and y, for x 9 in the propofed equation.

Now the
of
r,

values of x are fome of the divifbrsis

which
;

the term

left

when you fuppofe

fome of the

x-=>o

and the values of

di vifors of

+
;

p +
q

thejy's are
r,

and of

-p - - r,

refpectively.

And

thefe values are in arithme-

tical progreflion increaling

ference unit

becaufe x

by the common difx9 x 1 1 , are in

,

that progreflion.

And

it is

obvious the fame

to any equation of

reafcning

may be extended
So that
the refolution

whatever degree.

this

gives a general

method

for

of equations whofe

roots are commenfurable.

RULE.
"
Subftitute in place of the

unknown

quantity fuc~
i,

ceffively
1
,

the

terms

of

the progrejfion

o,

&c. and find

all the divifors

of the/urns

that refult,
progrejjions

then take out all the arithmetical

you can find among

thefe divifors,

whofe

: :

Chap.

6.

ALGEBRA.
is unit',

J95
and the va-

whofe common- difference

lues of

zvill be

among

the divifors dr'tfmg front

the fubftitutions of % == o that belong to ihefe

9

frogreJJionsJ

The

values of x will be

affir-

mative when the arithmetical progrefiion increafes,

but negative when

it

decreafes.

EXAMPLE.
60. Let
it

be required to find one of the

roots of the equation

Xs

io#+ 6=0,
X*
\Di cuifors.\Aritb.prog
i

The

operation

is

thus
deer.

Stlppojit.

Refillt.

*=

'

?
3

1,2,4
3

gives

x-i

C-r-i

Where
6,

the fuppofitions pf xz=zi 9

1
a-

give the quantity # 3

tf=0, #= ior+ 6 equal

to

4,

14

among whofe

divifors

arithmetical progrefiion, 4,

we find only one 3, 23 the term of

we

which oppofite
ing 3, and the
fubftituted
for

to the fuppofition of
feries decreasing,

x makes

the equation
its

which fucceeding one of

roots

muit be
try if
-{-

= o,

be3

vanifn
3.

Then

dividing, the equation

by x

3,

we

find

the roots of the (quadratic) quotient

xx
it

-f 2

=o
z

are 2

i \/2.
the

61. If

is

required to find the roots of

the equation

%x

46* 72^0,
O

operation will be thus

Suppf.

*9&

A Treatise
5*
II
I

of

Part

II.

<*
II

H
II

Co

O
T
1
1

GO
-*
\

^vj
to
*-c

to

^
^r
T

ft

x
to
w

O M
SI

to
s

to
s

oo <J On
**

oo
\*

GO

4*
*

>
s

M
-j

c>

o M

C\ Ol oc ON v* v vo oo
V*

hi

CXI

O
,

CO
to

M O M
to
fe? c3
^
ts>

4> v oo Q\ %*
^vj
to

O
to.
.

'

'

4> M GO
V*

4^

O
w
to

Sf

O
o

CN

HH

o
M
to

VD
to

CO

GO
-^

> n
*i

oo
4>

f
Si

GO

Oi

Of

thefe four arithmetical progreflions

hay-

ing their

common

difference equal

to unit, the
firft

Chap.
firft

7.

ALGEBRA.
all

197

= 9, the others give#z=-2> # = -3, which fucceed except x 3 fo that the three values of x are + 9, 2, 4.
gives #
;

x = ** 4

CHAR
Of

VIL

the Refolution of Equations by

finding the equations of a lower

degree that are their divifors.

62.^X^0

find the

roots of an equation

is

A
equations*

the fame thing as to find the fimple

by the multiplication of which into

it is

one another'

produced, or* to find the

it

Am-

ple equations that divide

der.
If fuch

without a remain*

fimple equations

cannot be found,

yet

if

we can

find the quadratic equations

is

from

which the propofed equation

produced,

we

may

difcover

its

roots afterwards

by the
Or,

refoluif nei-

tion of thefe quadratic equations.

ther thefe fimple equations nor thefe quadratic

equations can be found, yet, by finding a cub c

or biquadratic that
equation*

is

a divifor of the propofed

it

the folution

we may deprefs mere eafy.

lower, and

make

Now

198

-^

Treatise

g/'

Partll.

Now,
pofe that
?nx

in order to find

the Rules

by which
fhall fup-

thefe divifors

may

be difcovered y

we

n nx + mx* nx
mx
z

r fimple
"J

>are the I quadratic

s

-\-

rx

L cubic
,

divifors

of the propofed equation

the quotient
arifing

and

if

reprefent

by dividing the

propofed equation by that divifor, then

or,

EX EX EX
ic

mx
mx
mx
z

n nx nx
7

~T~
z

'<\

~T~

rx

r, s.

will reprefent the

it

propofed equation
that
fince

itfelf.

Where

is

plain,

is

the coefficient of the higheft

it

term of the

divifors,

muft be a

divifor

of

the coefficient of the higheft term of the pro-

pofed equation."
6$.

Next we

are to obferve, that, fuppofing

the equation has a fimple divifor

we
tity

fubftitute in the equation

n E x mx
mx

if

#,

in

place of #, any quantity, as a y then the quanthat will

refult

from

this

fubftitution will

neceffarily
fince,

have
this

ma n

for one of

in
n.

fubftitution,

ma

mx

its

divifors

becomes
arith-

If

we

fubftitute fuccefiively for x

<2,

metical progreffion,
quantities
tions, will

ey

any

2,

&c. the

that

will refult

from

thefe fubftitu-

have

among

their divifors

ma

Chap.

7.
#,

ALGEBRA.
%

199

ma ma ma
tical

me n ime

#,

which are

alfo

in

arithmediffe-

progreflion,

having their

common

rence equal to me.

If,

for example,

of

this

progreflion,

that refult have

metical progreflion
or,

changing the

1,0, among mn m n + n m,
1,
tz,

we

fubftitute for

x the terms

the quantities

their divifors the arith#,

\

figns,

#,

;;/.

Where
is n.

the difference of the terms is m, and the term belonging to the fuppofition of x o

64. It

is

manifeft therefore, that

when an
fubfti-

equation has any fimple divifor,

tute for x the progreflion
1,

o,

if

you

1,

there will

be found amongft the divifors of the fums that

refult

from

thefe fubftituticns, one arithmetical

at
leaf!:,

progreflion
will

whofe

common

difference

be unit or a divifor

of the coefficient of

the higheft term, and which will be the coeffici-

ent of x in the fimple divifor required

and
of

whofe term,
x

arifing

from

the

fuppofition

= o,

will

be n the other

member of

the fim-

ple divifor

mx

n.
this

From which
vering
any.

Rule

is

deduced for difco-

fuch

fimple

divifor,

when

there

is

RULft

200

Treatise

cf

Fart IL

RULE.
f
6

Subftitute for
ceffivdy the

in the prop of ed
i,

numbers

o,

equation fuc~

i.

Find

all the

divifors of the funis that refult

Jiituiion,

from

this fub-

and

take

out all

the arithmetical

progreffions
difference
efficient
is

you can find amongft. them^ zvhofi

uni>\

or

fame divifor of the

the
to

co^

of the higheft term of

equation*

Then fuppofe n equal

progreffion that arifes

that term of any one

the fuppofition

from

of

= o,

and

=
,

aid divifor the for ef

of the

coefficient

cf the higheft term of the equation^

is alfo

which

the difference of the terms of

this progreffion

fo

fihall

you have

mx

for the divifor required."

You may
is

find arithmetical progrefTions givwill not iiicceed

will
,

ing divifors that

any divifor,
it

but

if

there

be found thus by means of

thefe arithmetical progrefTions.

ficient

6$. If the equation propofed has the coefof its higheft term =: i, then it will be
T,

#==?
the

and the divifor

chapter,

will

be x

and the
after a

rule will coincide with that given in the end of

laft

which we demonflrated
5,

different

manner
that

for the divifor being

the value
progreffion
grifes

of x will be
is

+ #,

the term of the

a divifor

of the fum that

from fuppofing x

= o.

Of

this cafe

we

gave

Chap.

7.

ALGEBRA.
in the laft

201
;

and though it is eafy to reduce an equation whofe higheft term has a coefficient different from unit, to one
chapter

gave examples

where that
yet,

coefficient

fhall

be unit, by

without that reduction, the equation

this

30; may

be refolved by

rule, as in the following

66.

EXAMPLE. Suppofe Sx> i6x + ix

z

-j-

10

= o,

and that
x
,

it

is

required
is

to
:

find

the values of

the operation

thus

SuppoJ.

Refults.

Dit'i/ors.

Progr.

*==
xzz.

Sxi 26**4-11* + 10 <

C+

i> 31,

X -I

+10 t-35

3 5

2,5,10. 2

The
efficient

difference of the terms of the laft arithis

metical progrefTion

2, a divifor
l

of

8,

the co-

of the higheft term x

of the equan =. 5,

tion, therefore fuppofing

the divifor 2x

lows that

5 2X 5 =
is

= 2,

we
jt

try
fol-

-,

which fucceeding,
o, or x

2-i.

The
:

quotient

the quadratic 4**

3
*

3# 2
O

o,

whofe roots are

""""; 4 * and 3 ,

fo that the three roots

are 2

-,

of the propofed equation

V

^,

j 8 O 4

4I

The

other arithm^ticai

202

A
it

Treatise

of

Part IT.
,

metical progrefiion gives x

+2

for a divifor

but

does not fucceed..

If the propofed equation has

67.
divifor,

no fimple
has not

then

we

are

to enquire
itfelf

if

it

fome quadratic

divifor (if

is

an equation

of more than three dimenfions).

An
may

equation having the divifor

in

mx z
and

nx +
of
if

be exprefied, as

the

firft article

this

chapter, by
ffitute for

X mx

x any known quantity a y the fum that

r for one of
it's

will refult will

nx + have ma* na +
x

we

fub-

divifors
x.

and,

if

we

fubftitute

fucceflively for
2<?,

the progrefiion a, a

ey

%e

&c.

the funis that arife from this fubftitution will

have

ma 1

na
nx
~ $e\
"

nx a n x a^$e +
ie
-\-

+
e

r*
? ">

&c
amongft
their
divifors,

reflectively,
as in the lafi cafe,

Thefe terms are not now,

in

arithmetical progreflion

-,

but

if

you fub-

tracl

them from
e>

a
that
is

2e,

3^

the fquares of the terms a 9

&c. multiplied by

divifor of the higheft

tion,

term of the propofed equa-

from

m&

Chap.

7.

ALGEBRA.
-.

203

ma

mX mX mX
00

e)

\
3

a
a

^1%

r,

3^1

&c

the remainders,

71

72

X A e r, 2^ r* Xa r X Ze

>

&c

tical

progreflion,

having
e.

ma ^ ke in arithmetheir common diffe'

rence equal to n
If,

for

example, we fuppofe the afTumed pro-

greflion a, a
2,
1,

o,
1/77

in
4n
272

e,

2e 9 a

$e,

&c.

to be

1, the divifors will

-f- r,

be

m
5,7,0,772,

+
-J-

r,
r,

which fub tract ed from 4,

leave

72

r,

r,
r, r,

an arithmetical progrefis

fion,

whofe difference

+n

and whofe term o for x

is

arifing

from the

fubftitution of

r.

From which
tion",

it

follows, that

by

this

opera-

if

the propofed equation has a quadratic

elivifor,

you

will find an arithmetical progref-

fion

204

A
m

Treatise

of
r,

Partll.
the coit

fion that will determine to

efficient

you n and

being fuppofed.

known

fince

is

unit,

or a divifor of the coefficient of the higheft

term of the equation. Only you are to obferve, z that if the firft term mx of the quadratic divifor
is

negative,

then in order to obtain an

arithmetical
tracl,

progreiilon,

you are not

r>

m n+.r
428j

but add the divifors

9

~~-r y

m+ n+

4m m +

to

fubi\

to the terms

m> o, nu

The general Rule from what we have faid, is,

68.

therefore,

deduced

"

Subftitute

in
1,

the prcpofed equation for x the

o,

terms 2,

1,

&c.

fucceffively.
refult,

Find
add-

all the divifors

of the fums that

ing and fubtracling them

thefe

number's, 2,

1, o,

1,

from the fquares of &c. multiplied by

a numerical divifor of the higheft term of the

propofed equation, and take out all the arithmetical progreffions that can be found amongft
thefe

fums and

differences.

Let
"+

be

that

term in any progrefficn that

fubftitution of

arifes

from the
diffe-

x =: o, and

let

n be the

rence arifing
'

from fub trailing that term from

;

the preceding term in the progreffion

laftly,

let
y

be the forefaid divifor of the higheft term

then jhall

mx

+_ nx

be the divifor that

eught to be tried."

And

one or other of
the

Chap.

7.

ALGEBRA.

205

the divifors found in this manner will fucceed, if the propofed

dratic divifor.

equation has a qua-

6g.
dratic x 4

$x
thus

Suppofe,
z

for example,

yx

5* 6

the biquadra-

== o

is

pro-

pofed, which has no fimple divifor; then todif-

cover
ration

if it
is

has any quadratic divifor, the ope-

<

."".

..

Suppofi

2o6

A
H
II

Treatise
H
II

of

Part

II.

H
i!

H
II

O
C>
to

tO

+ 111
to

g^

CO

to

to

to

to

~
4^

'Spjtvnb^

Co

^r
Oi

1 w
V

Si

tw

5^

o
N>

to
X*

CO

Lo
O-i

^
CN.

8^

to
cc
X*

c 2

so

on

Ui

oo

>-<

CO

tO

Tk

Chap.

7.
firft
z

ALGE BR A.
arithmetical
2
;

207
z

%x the fecond gives x 2x the roots of the two both which fucceed, and # 2^ + 30, equations x $x 2 = + and the four roots z JL \ V 2
divifor x
-f 3
fo that
x

The

progreiTion gives the

o,

*vi

.*7.

2, are

of the propofed equation, the two

are impoflible.

lafl

of which
not
fuc-

The

divifors

which the other

arithmetical
ceed.

progreflions

give,

do

70. After the fame manner a Rule may be difcovered for finding the cubic divifors, or
thofe of higher

dimenfions,

of any propofed

equation.

Suppofe the cubic divifor to be mx*

+ rx
of
the
:

nx

j,

and by fuppofing x equal to the terms

it

arithmetical progrefiion,

will

be as

follows

Cubes of
Suppof.
Refults.

terms of
prcgr. x?n.

ift Differ.

id

Dijf.

""

gn+yi 7.7m n-\-ir Sm *== m x O m * = -I m n r

x=z

2
I

27m %m

6r

?n

vr\r

gn An n

y+

$n

Z**-f- s

r-\-s

s
i

+
a 4-

n n

r 2
r
2*1

r ZK

r+*j

Where

the

firft

differences are not themfelves in

arithmetical progreiTion, as in the lafl cafe, but

the differences of
ferences,

its

terms, or the fecond difprogreiTion,

are

in

arithmetical

the

common

2o8

^Treatise
difference being
is

of

PartlJL
is

common
The
fome
term.

2;?,

whence n

known.

quantity r

found in the column of the

s is

fecond difference, and

always to be affumed

divifor of the laft

term of the propofed

firft

equation, as

m is

of the coefficient of the

all

mx*

nx
it is

Whence
z

-\-rx

the coefficients of a divifor

.f,

with which

trial

is

to be

made, may be determined.

If
a divifor of four dimenfions that
is

re-

quired, by proceeding in like manner, you will

obtain a
ferences
is

feries

of differences whofe fecond

dif-

are in arithmetical progreffion.

five

If

it

a divifor of

dimenfions that

is

required,

you

will obtain, in the

fame manner, a prowill

greffion

whofe third differences

progreffion
-,

be in arith-

metical

and by obferving thefe

difcover rules
for dere*

progreffions,

you may

termining
quired.

the

coefficients

of

the

divifor

The
a

foundation of thefe Rules being, that,

<2,

if
9

an arithmetical progreffion

a 4-

e,

+ 2e

+ 3^,

&c.

is

affumed, the
will

firfl

differences of

their

fquares

be in arithmetical

progref-

fion; thofe differences being lae

lae+^e 7;
is

2ae+ $e\ &c. whofe common

difference

a
.

And

the fecond differences of their cubes, and

the third differences of their fourth powers are

likewife in arithmetical progreffion, as
is eafily

demonflrated.

s 7 i.

Chap.

7.

ALGEBRA.
But the fame
rules

209

71. Hitherto wc have only ihewn how to find the divifors of equations that involve but
one
letter*

ferve for disletters,

5

covering the divifors when there are two

if all the

terms have the fame dimenfions

the
letters

for,

"

by- fuppojing one of

equal to unit,

find the divifor by the preceding Rules^ and then by

co?npkating the dimenfions of the divifor^ fubftit til-

ing the letter again for unit, you will have the
divifor required"

Suppofe,

for

divifor of Sx z

i6ax*

example, you are to find the

-\-

11 a x

10a 3

= 0,
l

puttings
-|

1,

that quantity

becomes Sx

26*"

by

iix+io = 0;
to be 2x

ix

66.

whofe divifor was found,

;

now
it

multiply the term

5 by

+a

to bring

to the

fame dimenis

fions as

the other, and the divifor required

$a.

72. Befides the

method

hitherto explained

for finding the divifors of lower dimenfions that

may

divide the

propofed equation,

there

are
fol-

others that deferve to be confidered.

The
all

lowing

is

applicable to equations of
it

forts,

though we give
fions.

only for thofe of four dimen-

Let the biquadratic x4

it

px*+ rx +
qx z
,

=o

be the equation propofed

is

and

let

us fuppofe

the product of thefe

two quadratic equax*

tions,

^Treatise x mx + n == o Xx kx + = o
2io
z
1

of

Part IL

*?*

p-^ +
/

the terms of which will be equal, refpe&ively,

to the terms of the propofed equation.

In
the

this

equation,
sy

and n being

divifors

of

laft

term

(viz. I) as

we may confider one of them known and in order to find m or k,

;

we need

only compare the terms of this equa-

tion with the terms of the

relpedlively,

propofed equation

which gives,

i.

k+mz=ip.
mk
ml

2.

+ +n=
/

q.

3.

-\-nkz=z rs

4.

n!=zs.

Now

in order to find an equation that fhali

involve only k>

and known terms, take the

that arife

two values of

from the
will
find,

firft

and

third equations, and

you

m=p k

=
-

(becaufe n
ks

=
ji
p.

-r>

by

c~

quation the fourth)

= ==
'-,

whence

pl%

kl l

rl ks

and k
/

the quadratic x*
x*

kx
I

= + =
o.

rj
;

___

and

o becomes

~ ^p^ X x + =

Chap.

7.

ALGEBRA.
this to
/

ifi

To
lail

apply

practice,
all

you

mud

fubftij,

tute fuccefiively for

the divifors of

the

term of the propofed equation,

till

you
x

find one of

them fuch,

that x

~
pi

rl

+
a

can divide the propofed equation

remainder.

without

EXAMPLE.
6x -f- 20,r 2 34* y%. If the equation x* The divifors of 35 are o is propofed. 35 -Jl

!>

*>

5>

35

that arifes
/

y ou P ut will not fucceed.

S

if

tne quadratic
if

But

5>

then the equation #*

kx +

you fuppofe
1+

that

is,

x1

ii__J

* 4-

o becomes

6x2^

34x5
and gives the quotient x
z

2535
remainder,

^
a;

which divides the propofed equation without

=
}aft

4* + 7
s,

o.

" In
divifor

this operation
/,

it is

unnecefTary to try any

the
if

that exceeds the fquare root of

term of the propofed equation.'*

propofed equation
is

And,

the

literal,

" you need

term that are

only try thofe divifors of the laft

of two dimenfions."
If,

in
jz
is

any fuppofition of 7, the value of k$

,

viz.

becomes

a fraction, then that fup,

pofition

to be rejected, and another value of /

to be tried.

74,


212
74.

A Treatise
By comparing

of

PartH.

the fecond and fourth equa-

tions of the laft article,

you may obtain another

l
9

value of k.

= q mkz=j> fo that (m being equal to p k) -r~q pk-\- %

For n

and k z
k

pk-\-q = PI -7

Which givei
So
that

=
-

ip

J ^p

+ +
/

j.

the

quadratic divifor required becomes

x7

~ ipT J ^P
fame time
/

+ l + jXX+l-O.
when
/

This divifor muft be

at the

tried

y, and

=
cafe.

the former expreffiofi

not ferving in that

By

this

formula, divifors

may

be found whofe

fecond terms

may be

irrational.

How

the divifors of higher equations

may be

found, when they have any,

may be

underftood

from what has been

menfions.

faid

of thofe of four di-

Supple-

Chap.

7-

ALGEBRA,
to

%%%

Supplement

Chap. VIL

Of

the Reduction of Equations by

Surd

divifors.
fix,

AN
kLx't +

equation of four,

or

more dimenratio-

fions,

although
+
-f-

it

may admit of no
that
is %

nal divifor,

may have one

px
z

irrational.

As

the biquadratic #

+ qx +

rx

+ s = o,
by any

which we fuppofe
rational divifor,

to

be
yet,

irreducible

may

by adding a fquare

iklx

multiplied into fome quantity n+

-f-

z be compleated into a fquare x

ipx

In which

cafe

we

fhall
is

have x z

+ ip + ^= -/#
this

i^j\

X &+ 4
To
ner,

an d *

found by the refolution of

an affected quadratic equation.

reduce a biquadratic equation in

man*

we

have the following

RULE,
* If the biquadratic
is

#* -\-px z

+ qx
.

-\-

rx

+ S Oj
tai:<?

where p, q, r, s, reprefent the given coeffici.ip*=:, ents under their proper figns, pit q
r i*p =
(3,

* == f

And
(3
,

for

_/&/*

integer

common

divifor of

and z, /to
//"

0/ a fquare number,

and which

#7*r

or r

flii

number i muft be odd, and, di-

* Aritbmet* Uni-verf. pag. 264*

vided

214

A Treatise

of

Part

II.

vided by 4, leave the remainder unity.

likewife for k feme- divifor

Write

of

if

is

an

even
if

number*
is

or the half of or

an odd

divifor

odd*

o
let

if

(3

==

o.

Subtract 7
I.

from ~pk* and

the remainder be
try

For
s by

g^put
7iy

and

if dividing
is

nl

the root of the quotient

if
it
is,

rational

and equal

to I,

add nk z x z

inklx

to

both fides of the equation , and extracting the

root

you jhall have x z

ipx

^ z=
nx

n^

kx

/.

EXAMPLE
Let the equation propofed be

I.

,v

-f

17

= 0,
we
and
a

and

becaufe'jp

fhall

have

= o, ^ = 0, r = 12, s=*-ij, 12, = 17. And a :0,

|3

(3

2^, that

is

12

and 34,
it

having only 2 for

;;

common

divifor,

muft be

=
*2

2.

Again,

=
for
/

6? whole

divifors 1, 2, 3, 6, are to be fuck,

cefiively

put for

and

k\

3,

~,
equal to

1,

refpectively.

But

that

is

is

^,

and

and 6

/.

And when

the even divifors 2

are fubftituted for k,

^becomes 4 and

36,

and

Chap.

7.

ALGEBRA.
s

215
is

and fible by n
fet afide.
is
1

being an odd number,

2).

not divi-

(=

Wherefore
1

and 6 are to be
for k,

But when

which numbers can

or 9, and

and 3 are written

is
1

<^

98 reflectively be' divided by 2, and the

3 or

roots of the quotients extracted, being

3, coincides 7; but only one of them, viz. with /. I put therefore k~ 1, Thz 1, 3, J3>

and

and
tiklx*

adding

to

both
id\
?

fides
is?
z

of
2.v~

the

equation

inklx

there refults x*

+ + 2# +

tnat
1

== 2x
z

i2# + 18,
1

extracting the root of each, x

-\- 1

2# + 8, and '+ V^ x * 3laft,

And

again, extracting the root of this

the

four values of x9
the figns, are

according to the varieties in

Wz + s/S'S ^ 7T/*-V3^'-|i W ++J 3^2 ~, 4-/2 J---3^2 ~, J \ W2 V

1
l

being the roots of # 4 -[tion at

fir ft

1 2.v

17 == o, the equaII.
z

propofed.

6x 58** 11 = and writing 6, 58, 114, 11 refpe&ively, we have 67 = for 315=3(3, and 1133^=:^. The numbers 315 and i522, have but and
Let the equation be x*
0,
^>, q, r, j,

EXAMPLE

1 14a?

a,

(3

2<,

that

is

one

common

divifor 3, that

is

= 3.

And

the

divifors

216
divifors of

^f

Treatise

of

Part

II.

105 == Wherefore
I

are 3, 5, 7, 15, 21, 35,

firft

and 105. 3

make
1

trial

with
it

= ,

and dividing

and

or

05,

by

get

the quotient

ipk-=
*

3X35
1

ti

%$,

this

fubtracted
half, 13,
>

from
ought
.

leaves 26,

whofe
,

be equal to
<

/.

But

+*

or

67 4- 27 s

that

is,

20
I

is

equal to Q-, and

= 411,
;

which
There-

is

indeed divifible by n

=3

but the root of

with 5

the quotient 137 cannot be extracted.

fore
reject the divifor 3,

and

try

= k\
is

by which dividing

1,

and

this

= 105, the quotient taken from ~pk = 3 X 5> leaves

i$

2/.

At

the fame time,

(=
or 16

j ==

^.^21^4,
divifible

And-^-j,

11,

is

by

that that

is,.

3,

and the root of the quotient 9, coincides with /. Whence I conclude

,

putting/
fides

3,

=5, 5L 4?
the
a
,

= 35
the

adding

to

both
-j-

of

equation
is,
it

quantity

nk % x z

inklx -f */

that

j$x z
will

-{-

90*

+27,

and extracting the

roots,

be
or

^+
x%

+ ^>= s/n x kx^\ 3* + 4 = V3 X ^+#

~px

9;^ 15** *7a?+9==o writing 9,+i5> +

In like

EXAMPLE

III.

manner

in the equation # 4

-J-

2 7>

9>

fQf

, ,

Chap.

7.
J,

For pi q, r,

there refult

ALGEBRA. 217 5^, 504 =

r= p,
divifors

2-^ = .
that
is,

The common

of

|3

and 2,

of ijp and

-^

are 3, 5, 9, 15, 27, 45,

135; but 9 is a fquare, and 3, 15, 27, 135 divided by 4 do not leave unity for a remainder,
as
is

required

when p

is

an odd number.

Set-

ting thefe afide there remain only 5 and

45

to

be tried for

n.

Fifft let n

of the odd divifors of

= 5, =

and the halves

5-,

that

is,

2' 2
,
,

are to be tried for k.

If= 2
a

the quotient

of
leaves 18

divided by k, taken from

^pkor
==

2
j

2-,

2/:

and
is 1
,

^ *==
/

^( = fI )
divifible

but the root of the

be

quotient

5>

which

by

5,

which mould

= 9,

is

imaginary.

Put next k

and
r

the quotient of
J.
is

divided by , or of

by

2'

-.

This fubtracted from ~pk r

.

4
0

leaves

nothing,
j

that
3>

is,

/=o.

Again,
*

^ (= 2-^- =
/

and C*

and

(= ^Z

= a.
tnat
z

From which
,

coincidence

infer that

= 5, k =
+ '%
I

== o,
2

and adding

k& 2 #*+ 2^fe

the equation,

is?

find x

4-^ + 3 = ^5 X

V*

10 both fides of

ix.

P 4

Literal

2*8

A Treatise
may
be

of
treated

Part IL

Literal equations

much

in

the

fame way.
will

fame Rule
for the

And, .if you put n ==_ i, the give you the rational divifor of
if it
l

a biquadratic equation,

equation x* x

admits of one.
-\-

-f 5*

putting

wni
is

find

=
+

i2x 6="o, = and

,

Thus

the equation $* 3;; -f 3

reduced to the two quadratics an d x z 2x 0. 2

When
all for

the divifor s of

are fo

many

that

it

would be troublefome
k 9 their

make trial with them number may be reduced by findto

ing

all

the divifors of us
half

4-r

z
.

For to one of
the

thefe, or to its

when odd,

number j^

niuft be equal.

The ground

of

this

Rule

is

as follows.
1

If a biquadratic equation

x^-^px -^ qx z -\- rx
are the given cois

+ = o,
j

in

which p y
their

q, r, j,

efficients v/ith

figns,

and the equation

;

fuppofed clear of fractions and furds

if this

equation can be compleated into a fquare, in

the

manner already
qx
z

defcribed,
s
z

we

fhall
-f-

have
z

x* -f- px*
z

+ + rx + -f nk x -f 2nklx nl == nk x + 2nklx + nfc = x + \px -f h that x + px + q + nt x x + r+ inkl x x + + nl == x + px> + 2^+ip' X x + pgix x + g*

z z
z
,

is,
l

And

comparing the terms, we get

1.

thefe three

equations,
2

+ vk* = 2Q+ip%

Chap,

7,

ALGEBRA.
2. r
3.

219

+ 2nkl p^ s+ **=* 3?*

which there being four unknown quantities, they can be found only by trial.
in

The

values of j^, taken from the

firft

and

fe-

cond equations and made equal

1

to each other,

give n

-If-

= 77_2 p (writing, as in the Rule, a and r 4op = =

1

(3)

-.

Whence,
found,
it

if

the quantities

tz,

k>

/,

^,
2/,

are to be
divifor

follows, (i.)

That n being a

of

(3,

giving the quote k

ipk.

(2

k will be a

divifor of

giving the quote ~pk

2/
I

and
be

that fubtracting this quote from ~pk 9

will

equal to half the remainder.

.)

In the

firft

equation
I*

we had i^==
0#
)

and, from the third,

__

g*

^ ^
j

Becaufe

^~ > + \nt and

-,

(if

= _ ^)
2 by k z

_.

-^

1\
it

that

is,

divides

x <*+ink z

And
/,

if

the

feveral values of the quantities , ,

j^, anfwer
is
,

to thofe conditions, or coincide,

a proof

that they have been rightly afTumed

and that
quantity

adding
n

to

the
it

given
will

equation

the

kx

+ /)\
%

be compleated into the

It

fquare x

+ip+<$\\_

220
It

A
was

Treatise

of

Part II.

^ will always be fome divifor For us =* cx.^ /% and taking 2 from both 4r ip*p2>nkl -f s*7% feeing
faid that

of

ocs

ir\

the remainder

fgpkl *=
It
is

a^ 2^x nl*ip $!; + p$nkl,

2 %
;

a^ a + W X
the thing
is

/*

i? ^ +
a

has

i^in every term

manifeft.

needlefs to be particular as to the feveral

limitations in the Rule, feeing they follow eafily

from the algebraical exprefiions of the quan-

tities.

You

are not, for inftance, if

ft

you feek

a furd divifor, to take

if

a fquare number, for

is

a fquare number, */n

kx

+ I would

be

rational.
v

Or

if

is

a multiple of a fquare, as

% then, at

leaft,

mxkx^l would be rav.

tional,

and n would be deprefted to

Let us examine one cafe, when p is even and r cdd \ and by the Rule n muft be an odd number, a multiple of 4 more unity.
1
.

Seeing
(3

numbers
is

fop, or (3 r |3 ?, of the ip and fop one muft be even and the

other odd, that their

fum

may

be odd.

If

(3

odd,

its

divifor

n muft be odd likewife.

Suppofe p to be even, then fop,

quently \p and a are both odd.

2
n

= / f* will
divifor
is

and confeBut if a is odd, be half an odd number, and

half an

its

odd.
is

In this cafe, *
let it

odd number.

For

be an integer,

^^wiil be an even num/,

ber.

But

if

j^is an integer, fo muft

becaufe


A L G EB R
i^f
5

Chap.
% s 4- nl

7.

A.
be even.

221

and 2^/

mud

And

+
2.

2/ (an odd number)

= pg^ an even numin general,

cc

ber,

which

is

abfurd.

Let TV reprefent any number

,

an odd number
is

then I fay,

number
that
is,

a multiple of fouri
1.

/= 4-/V +
is

every odd more or lefs unity" u The fquare of an odd

<c

number

4iV+

1," and

if

from fuch a fquare

there be taken any multiple of 4, the remainder,

if greater

than unity, will be 47V-f-i/'

it

Hence

follows that n 3= 4.ZV

ing nlx -=:6g

1.

For

fee-

becaufe

and i^are the halves

of odd numbers, we have, according to the preas Iz nX I1

fent notation,
==*

or without the

common n X 4N-\-i
n

4 denominator n

I2 ==
15

I*

4-r,

that

is,

4iV~fit is

and confequently,

= 4iV +
In
like

For

not 47V

but

4N+

that can give the product 4JV-J-I.

manner the other limitations may be determined. We mall add only this remark more.

That
cafe

if

o,

neceffarily
9

(3

=o
(3

for in this

p^-pgj^o.
true,

^_=fa and p = r whence

this
is

(=r

i&p)

But the converfe of

although
?i

not univerfally

it

is

for

moil

part.

For

(3

be-

ing = x k x ipk
quantity, if tp

2/,

although k be a real
(3

= 2/or l=fpk,

will vanifh.

In this cafe

we may

ufe the following

RULE

2,22

^Treatise

of

Part II

RULE
" Of
that
is

I.

the

compound

divifors of 2< take

one

a multiple of a fquare number, but itfelf

if

not a fquare* as nk\ and try

the quotient

$ &.

-J~

f nk %

ff-

-,

if this

happens put

Thus
-|- 1

in the equation

#4

= o,
I

where

= 38, = o,
(3

-J-

2x 3

37** 38* = 720:

2<

whofe divifors
only
trying

are

many, but by
leaft to the

infpeclion

can reject the

with

purpofe, and
the quotient

45

x
'

3-,

H ~ 16 = 38

-j- 41.

-. 2

Which gives &

RULE
* When
fome
P

II.

= o,

and likewife
is

= o,
/.'*

being

divifor of 2C, if

a fquare

num-

ber, the root of this fquare mall be

As
it

in the equation
(3

#4

+ 2#*+ 6x +$x 5 = 0,
z

being a

= 5, = o,
that
is,

2^ ==
/

taking

0=5,

^'=f,
^+
*

= !,

=f

And

+ I = ^5

Xf
What-

Chap.

8.

ALGE.BR A.
has been faid

223

What
higher

may

lead to the inven-

tion or demonftration of fimilar Rules for the

equations

of even dimenfion&,

if

any

one pleafes to take the trouble.

CHAP.
Of
kind.
75.

VIII.

the Refolution of Equations by

Cardans Rule, and

others of that

TT 7 E
VV

now

proceed to fhew

how an
only
I.
;

expreflion of the root of an equafhall

tion can

be obtained that

involve

known

quantities.

In Chapter 11.

Part

we
and
any

fhewed how to refolve Jimple equations

in Chapter

13.

we fhewed how

to refolve

quadratic equation, by adding to the fide of the

equation that involves the

unknown
it

quantity,

what was

neceffary to

make

compleat fquare,
fides.

and then extracting the fquare root on both

In 27 of this Part, v/e gave another method of refolving quadratic equations, by taking away
the

fecond term
9

where

it

appeared that
z

if

px + q=:0

x-=L^p_sJ^p

q.

76. The fecond term can be taken away out of any cubic equation, by 25 fo that they all
-,

may

be reduced to

this

form, # *
3

+ qx + r = o.
Let

224

^Treatise of PartIL Let us fuppofe that # = # + ; andx + qx + r = a + ^cfb + 3^ + + qx + r =z a + X^+ ^ + ^ + ^ + ^ = ^ + 3^^ + ^ + ^ + r (by fuppofing 3^ == q) = + + r = o.
l

*z

But

-, and 3

=r*

^-r,

and confer

quently,**

j^j + rsio;
m%
9

or,

+ r^+^.

Suppofe a %

and you have % %

+ rz = =
__

which

is

a quadratic whofe refolution gives

= ir s/t^ +
3

$j

^
and

and

jx^'t ir N /#r*+^;

^=+i= i=v _| r ^ ir .^ +y 3
;

27

in

which ex-

preflions there are only

known

quantities.
littfe

77.

The

values of #

may

be found a

differently, thus

Since a>

= fr + ^ f + -,
a

it

follows*

that **

+ r = + ir +

,/

r*

+A

and

Chap.
*

8.

ALGEBRA.
!

225

(=

= irTVtr
r)
3

-f 5 fothat 27

_V ir + y-Jr' + l^i 27

and#(s= + $;s5

which gives but one value of

in the value of a the furd
fitive,
is
it is

#, becaufe

when,
is

^/

Jr* -{-

po-

negative in the value of , and there

only the difference of this fign in their va-

lues.

So that we may conclude

= J-ir+Jir-++/-ir-Jv>+. T 27 27
'

* The values of x may be difcovered 79. without exterminating the fecond term.

Any
form,

cubic equation

may be reduced

to this

&

3px z

3#H %r + 3P * t + 3??
%
l

which, by fuppofing
to

*
is

sqz ir =
wanting.

#=z-J-p,
o,
in

will

be reduced

which the fecond

laft article,

term

But by the
j
l%

fince

# *
3

z 3<l

2r =

follows that

* Vid. Phil. Traxif. 309.

fuppofc

226

.4

Treatise
root

of
of the

Partli
binomial

fuppofe that the cubic

r J- vV z

im.
:

q> is

m -\-

\/n) == m-\-s/n-\-m
y

s/H

And
-j-

fince

x~z-\*p

it

follows that

x =r p

2m.

is

79. But as the fquare root of any quantity twofold^ " the cube root is threefold" and

can be expreffed three different ways.

Suppofe the cube root of unit is required, 3 i, or y - 1 o, then fince unit and let j itfelf is a cube root of 1 ? one of the values of
3

is

s,

fo that the

equation^

1=0
o,

fhall divide

the

fiift

equation/*
1

and the quotient

y^J^yJ^

= o refolved,

gives y

2
"" T

fo that the three expreilions of y/iarei,

"^

~3
,

and
root

of

"

And,

in general,

the

cube
or

"

->

any

quantity
'~"
T

A*

may be A>
~~ 3

X -4

or

x -^i

fo that the

cube root of the binomial r

-f

vV z
or

q*

may be

^ ~j- *J n

as
i

we fuppofed above,
I

^3
a
-j-

X + ^j
i.
a.

-f-

*/

or

X m

And

hence we have three exprefilons for x, viz.

x =.

2,
f
-

g.

# = ^ m + ^ 3 x p ^ v
-=i

37* ;

and

Chap.
and

8.

ALGEBRA*

227
propofed

thefe give the three roots of the

cubic equation.

EXAMPLE
So. Let
it

be required to find the roots of

the equation x*

2x

+ 4*'# 42 = o.
of
this

Comparing

the

coefficients

equation

with thofe of the general equation

+ 3P
i.

S "

4-

2pz=
z

12,

fo that

.........

m
z=z

f \ = i
3

y u

find >

p 4,

2*.

3^

3?(=4S 3^ =4i
r*

2 |*

the

and confequently,
and r

-=Now
to
is

-f-

Vr

= 3 + y i%
3

cube root of

this

binomial
*

found

bt

4- .-J

4 (=

+ 00
1

Whence,

+ 2^3 z= 4 2 =2. 2*.x=p m V 3 = 4 + -/4=5-2 = 3. y.X=p + V = 5 + 2 = 7.

!*.#=/>
?;;

3;;

So

that the three roots of the propofed equation

are 2, 3, 7.

You may
root of 3

find other

two

expreflions of the cube

J ^,

befides
Part

I.

-f^/

* Se&ion 131.

228
viz.

^Treatise
+ J v 2
-, and
12'

of

Part

II.

i J 21 .2 v
12
II.

but

thefe fubftituted for

m -f v7 #

give the fame va-

lues for x\ as are already found.

In the equation # 3

you

find
i

./>

EXAMPLE + 15^4- 84^ 100 = = ^=135, and

o,
5,
3,

r _|_\/ r
is

~ 135 -|-v
fo that

18252, whofe cube root

+ Vi 2

fj

( ^ -{- 2^y

The
,

other two values of x, viz.

are

-5 + 6=1. 8 + V-36,

^6, are impoffible.

After the fame manner, you will find that the

roots of the equation x
3

-{-

66x-\- 660

= 0,

15,
if

7\/$-

The Rule by which we may

difcover

any of the roots of an equation are

be demonstrated afterwards.
roots of biquadratic equations

impeffibki mail

82.

The

may

be found by reducing them to cubes, thus.

Let the fecond term be taken away by the

Rule given

in Chap. 3.

And
s

let

the equation

that refults be

* 4 * 4- <F 2 4~ rx 4"

And

let

us fuppofe this biquadratic to be the

product of thefe two quadratic equations,

x% x
%

-\-f- ex

/=

ex -f- g

=o

Where

Chap.

8.
e
is

ALGEBRA,
;

229
becaufe

Where
tions

the coefficient of x in both equa-

but affected with contrary figns

the fecond term
is

when

wanting

in

an equa-

tion, the

equal to the

fum of the affirmative fum of the negative.

roots

muft be

Compare now
a"bove product,

the propofed equation "with the

'

and the refpective terms put

equal to each other will give

eg

f -j-

ef = fg ~ g = q 4" *% and
r,
s.

f -\- g
it

Whence
g

q9 follows, that

?=

fdz\
=zq

and confe-

quently

and g

=.

/+ + _/(=- %g) 9 + +7
<*
,

+ t*'+^
you

the fame way,

will

find,

by

fubtraction,
I

&c.

f=

+ ^-7
,

and
and

/ X g (= s)
have

= - x + 2^* + ^ _ r

1
.

multiplying by 4^% and ranging the terms, you

this equation,

+
*

2^ 4

+ ^ 4^
and
it

=
r
%

o.

Suppofe
2*

=: j,
2

457

r = o,

becomes j 3

-{-

iqf

-{-

a cubic equation

whofe roots
articles.

are to be difcovered

by the preceding y being found, y =

e
z

Then

the values of

their fquare
e9

roots will give e (n\nce

and having

Q^2

you

230
you
will

A
find

Treatise
/
*

of

Part II*
the

and

g from
-

equations

/=
x
r

Y~L
ex -f

g
you

~-

Laftly, ex-

trading the roots of the equations x z

+ ex +/= o,
of
;

->

will find the four roots

the biquadratic

x=z

-^jtV-i** /,
Or
if
-

x^qx z +

rx

or,

+ o for either x~ + ie'Jie* g.

s

83.

you want to

find

the roots of

the biquadratic without taking

away the fecond

term

fuppofe

it

to be of this form,

4PS*

+mS' + fl
will
*
.

and the values of x

be
I %

xp as//f+q a
.

2r")
\

>,
x

where

and x
a%
is

p r a s/p +q'a +J
-

equal to the root of the cubic,

'*->}/ JTl' -" =

The
article,

demonftration
as the

is

deduced from the

laft

78 th

is

from the preceding.

H A P.

Chap.

9,

ALGEBRA.

231

CHAR
Of
meral Equations by
84.

IX.

the Methods by which you

may

approximate to the roots of nutheir limits.

is

TT7HEN,

any equation
firft

propofed
the
li-

VV
are required,
T

to be refolved,

find

mits of the roots (by Chap. 5.) as for example, z if the roots of the equation x i6x -j- 55 -=zo

you
;

find the limits are o, 8,

is,

and

7>

by 48
8,

that

the lead root

is

between

and

and the

greater!:

between

and 17.
I

In order to find the

sider that if I fubftitute

firft

of the roots,
2

con-

for x in *

16^ + 55,
and confe-

the

refult

is

pofitive,

viz, -\-

5^

quently any number betwixt o and 8 that gives

a pofitive refult, muft be
lefs

than the lead root,

refult,

and any number that gives a negative

muft be
pofing *

greater.
is,

Since o and 8 are the limits, I

try 4, that

the
a

mean betwixt them, and

is

fup-

= * 16*4-55 = 16 64 + 55 = 7,
4,
I

from which
than
4.

conclude that the root

greater

between 4 and
fubftituting
it

So that now we have the root limited 8. Therefore I next try 6, and
for

96+55 5\
1

x we find x

i6x

-\-

$5

= 36

which
is

refult

being negative,

conclude that 6

greater than the root reis

quired, which therefore

limited

now between 4

0^3

mi

232
and
6.

A
And
in

Treatise
I

of

Part

II.

fubftituting 5, the
find

55 and confequently 5 is the lead root of the equation. After the fame

25 80 4- 55 o
manner you
root of that equation.
85.

them

place of #,

i6x 4-

mean between

will difcover

1 1

to be the greater!:

Thus by diminifhing
lefTer limit,

the- greater, or

increafmg the

you may difcover the

true root when it is a commenfurable quantity. But by proceeding after this manner, when you have two limits, the one greater than the root, the other lefTer, that differ from one another but by unit, then you may conclude the root is

inccmmenfurahk.

We
pofe x

may however, by
is

continuing the opeit.

ration in fractions, approximate to

equation propofed

6x-\is

As
if

if

the

2,
is

the refult negative,

4 124-7 = ^
*

=0,

we

fup-

which being
$
x

and the fuppofition of

refult,
2.
--= 1

o
=
1 %

giving a pofitive
betwixt o and
a whence #

it

follows that

the root

Next we fuppofe

6.v

+7

6~)-7r--:+2,
is
it,

which being
twixt
rable.
1

pofitive,
2,

we

infer the root

be-

and

and confequently incommenfu-

In order to approximate to

we fuppofe

= if,

and find x %

6x

-j-

= -t 9 + 7 = ~
!

and

this refult

being pohtive, we infer the root

rnuft be betwixt 2

and i|.

And

therefore

we
try

Chap.
try 14,

9.

ALGEBRA.
*
6x

233
7

and find

= t? 12+
whic h
is

= 3tV 10A + 7 = ~^
tive
,

nega-

fo that

we conclude

the root to be betwixt

1^ and ii> which giving

the root
therefore
is

And

therefore

we

try

next

1-f-,

alfo a negative refult,

we conclude
i-f.

betwixt \\ (or 14) and

We

try

1^, and

the refult being pofitive,

we

conclude that the root muft be betwixt 1-^ and

1

44, and therefore

.

is

nearly \\ 9 TT'
eafily

S6.

Or you may approximate more

by transforming the equation propofed into another whofe roots (hall be equal to 10, 100, or

1000 times the

and taking the
portion.

roots or the former (by 29.)

limits greater in the
is

fame pro,

This transformation

eafy

for

you

are only to multiply the fecond term by 10, 100,

or

coo, the third term by their fquares, the

fourth by their cubes, &V.

lait

The

equation of the
x

example

is

thus transformed into x

6oox
li-

+
we

70000

== o,

whofe roots are 100 times the

Proceeding

roots of the propofed- equation, and whofe

mits are 100 and 200.

try

22500

6oox + 70000 = 90000 + 70000 = 2500, fo that 150

150, and find x
%
:

as before,

is lefs than the root. You next try 175, which giving a negative refult muft be greater than

the

root

and thus proceeding you and 159

:

find

the

root to be betwixt 158

Q^4

from which you

234
ycu
tion

A
x
z

Treatise

of

Part

II.

infer that the lead root

of the propofed equa-

6x 4~ 7 i s betwixt 1.58 and 1.59, being the hundredth part of the root of bOCX ~j- 7COOO == o. Xz

50 =
~f-

l 8 J. If the cubic equation x

iqx

-f
z

63%
30.V

is

propofed to be refolved, the equa-

tion of the limits will be (by

48)

3x

63
>>

=0,

or x -~
z

io.v -f

21

z=l

o, whofe roots

are 3

of x

',

and by fubftituting o for x the value

ijtf*

63x

50

is

negative, and

by

fubftituting. 3 for x, that quantity

fitive,
it

becomes poand

gives

it

negative, and x =- 2 gives

is

pofitiye, fo that the root

between

2,

and

therefore

incommen (arable.

You may

pro-

ceed as in the foregoing examples to approximate

to the root.

But there are other methods bv

that

which you may do

which we proceed
88.

more

eaiiiy

and readily*

to explain.

When
lefs
it is

you have difcovered the value of

than an unit (as in this example,
a little

the root to

you know

above

1)

fuppofe the diffe-

rence betwixt

its real

value and the number that

to
it,

you have found nearly equal

fented

to be repre-

by/

as in this

example.

Let

x~

-f-/.

Subftitute this value for x in this equation, thus 3

i5* =-r 6 f 6 4- 6 x = 5 = 5
:

*'=

?.-f 15 3/
3 -\-

3/+
l

3f z 5f

+P.

' i5x +63x o = Ti + 3 6f-i 2f~ +/ 5

z

= o.
Now

Chap.

9-

ALGEBRA.
/
is

235
than unit,
its

Now
powers
ximation
terms,
ib that

becaufe

fuppofed

lefs

/% /
-,

may be

neglected in this appro-

fo

that alTuming only the

two

firft

we have

1+36/= o,

or/ =

= .027

x will be nearly 1.027.

a nearer value of x

You may have

it

fidering, that feeing

+ $6j ~
r

by con-

2/* -f f*

= o,
/?

follows

that

/=
nearly
'
1

36-^/+/*
;

(b

fubftituti "g

= 6 - I2x t^+t^Xt3 = -7 = .02802. r * 46225

o
;

I2 9 6

for

89. But

the value

of

/ may

be corrected

and -determined more accurately by fuppofing

to be the difference betwixt that
its real

value,

and
it.

which we

laft

found nearly equal to

So

that

/=

.02803

+ g.
it

Then by

fubftitut-

ing this value for

in the equation
will ftand as follows,

if"

+ 36/ 1=0,
+3^

y^zz 0.0000220226 + 0.002357^+0.0849^4-^3

z \2f ZZ- .OO942816 $6f =1 I.OO9OO

Q.6j2J2g I2

1=

--0.0003261374+35.329637^11.9195^+^-0.

Of
reft,

which the

firft

two terms, neglecting the

give 35.329637

X = 0.0003261374,
o. 000009 2 3
1

and
.

= ~r: =
000^261374
.

27.

So

that

/=z O.02803923
which
hat
is

02803923127; very near the true root of the equation

I

27; and #

-]-/:=

1.

was propofed.

236
If
(till

Treatise

of

PartlX.
is

a greater degree of exactnefs

re-

quired,

fuppofe h equal to the difference be-

twixt the true value of g, and that

correct the value of g.

ready found, and proceeding as above you

we have almay

90. For another example

3 to be refolved be x

ix 5
2

let

the equation

== o,

and by

fome of the preceding methods you difcover one

of the roots to be between 2 and
fore
this
3.

There-

you fuppofe x == value for it, you

2X
5

+ /,

and fubftituting

find

f
z=z

=4
=5

2/

I*
10/=
1

= -1 + 10/+ 6/*+/From which we

o.i. find

that

nearly,

os

Then
o. 1

to correct this value, g,

we
3

fup-

pofe

/=
/*

+
1.

and find

= 6/* = 10/ =

0.001+ 0.03^
0.06

1.2

+ 0.3^ + g + 6. g
x

+10.

m
.

0.061

+11.23^+6.3^+^,
==

io that g A

= 0.061 11.23
its

Then by may correct

0.0054. ^ fuppofing g=. .0054 + h

value, and

you

you

will find that the

*OQt required

is

nearly 2.09455147.

Chap.

9.

ALGEBRA.
you may

237

91. It is not only one root of an equation that can be obtained by this method, but, by

making

ufe of the other limits,

difco-

ver the. other roots in the fame manner.

equation of 87, x 1
has for
its

15^+ 63^ 50 =
be nearly 1.028039.

The
o,

limits o, 3, 7, 50.
leaft root to

We have already
If

found the
it
is

required to find the middle root,

you proits

ceed in the fame manner to determine

limits to

neareft
for

be 6 and 7

for

6 fubftituted

gives a pofitive, and 7 a negative refult.

fore

There-

you may fuppofe x

= 6 -f-/,

and by fub-

ftituting this value for

x in that equation, you

fo that

find

p 4- 3/* 9/-f 4 = 0,
Or
fince

/=
(by fab*

nearly.

/=

--, it is

ftituting

for /;

/=

^
Which

= |^,
value
articles.

whence
ftill

6 + |-^ nearly.
as in the

may

be corrected
the fame

preceding

After

manner you may approximate to the

value of the higheft root of the equation.

92.

"

In all thefe operations,

you

will

apif

proximate fooner to the value of the root,

you take the three

laft

terms of the equation,

and extract the root of the quadratic equation

confiding of thefe three terms."

Thus,

in SZ> infcead

of the equation

i2/

of the two
a

~f-

36/ =
1

laft

terms
o, if

you

238
you take the

Treatise
l

of

Part II.

three lafl and extract the root of the

quadratic 12/*

Z^f r\- =
is

> y ou will find

=2 .02^031, which

much

nearer the true

value

than what
1

you difcover by fuppofing^

36/
It
is

o.
this

obvious that

method extends

to all

equations.

93.

" By afTuming

equations affected with

general coefficients, you may, by this method,

deduce General Rules or Theorems for approximating to the roots of propofed equations of

whatever degree.

53

Let/
tion

pf

-f-

qf

in

r == o reprefent the equa-

by which the
is

fraction

is

to

be deter-

mined, which
fubtracted

to be added to the limit, or

it,

from

order to have the near

value of

%;

Then qf

r=o
,

will

give/=

for

But
/,

fince

=
-

Fpf + f
Theorem
r

by fubftitutine J
for finding

we have

this

nearly,

mix,

/=

fx

After the fame manner,

if it is a

biquadratic,

by which /is

rf~\-s =
/
==

to be determined, o, then
^

as/4

pp-j-qf
we

2,

/being very
is

little,

fhall

have

which value

corrected by confidering


Chap.
9.

ALGEBRA.
/=
r

239
(by fubftitut-

fidering that

pf

,_p

ing

for

/)

k whence we have

tjiis

Theorem

for all biquadratic equations,

94. Other Theorems

may

be deduced by

alTuming the three terms of the equation, and

extracting the rcot of the quadratic which they

form.

Thus, to
1

find the value of

P Pf + 8/ r
be very
little,

in the equation

wnere / is fuppofed to
firft

we

neglect the

term/
z

trad the root of the quadratic pf

or of

qf

and exr

-\-

= o,

~ X /+ ~ =
value of

and we find

But
poring
in

this
it

/ may
z

be corrected by fup-

equal to w, and fubftituting m> for

3

the equation/
z

givew 3
== o
;

qf pf +qf r=o,andjp/ qf+r m

z

pf

o,

which

will
l

the refolution of

which quadratic equation

.

gives

^ = q+'Vq
%
?
-

d.pr-\-A.pm*

very near the

true value

of/.

After the fame manner you

may

find like

Theorems
tions, or

for

the roots of biquadratic equa-

of equations of any dimenfion whatever.

9 $.

240
n

^Treatise
^

of

Part II.

95. In general, let #-{-# z -[- 2**-*-} rx 3 +, &c. -j- ^f =.0 reprefent an equation of
any dimenfions
n,

where

is

fuppofed to reprethe equation.

fent the abfolute

known term of

Let k

reprefent the limit next lefs than any of

the roots, and fuppofing x

= k + f>

fubftitute

the powers of k
#,

+/

inftead of the

powers of

k+fn + p X k+f n ~* n &fa. + = o, f X F+/-^4- r X +7

and there
will arife
""~

+
or

5>

^tf

by

involution, difpofmg the terms according to

the dimenfions of
** **

/
*
?**

^
^*

*
1

a
I

**

*r-

+ ^
X
l 00
*
it

+
**

+
X
Hi 1 ?*

f M

+
^
a

X
1
fc3
I

'1
a
1

?*->

a
i

X
\

i X

+
X
oa
1

+
t-Cs

+ ^
X
1

""S

X
'1 to

X
ll A
?
to

SI
X
1

X
I

1
!

#\
!
V*

N>

K
1

^ a
1

1-*^

is*>

^>
t*

^ ^
<

-fc

+
$

Q ^

+
<
^>
1!

<S
*"

*>

+.
tw

where

Chap.

9.

ALGEBRA.
all

241

where neglecting
rlt

the powers of
find

after the

two terms, you

?*-

^s
1 1

x
11 as **~
1

+ + 3 + M X
1

j.

St

*.

^o-

X
to.
it
1

*3

it
1

?l
|

II X

+
*>C5

i\
*s>.

+
*Q

^
1

a
1 1

a
1

X
I
SAtJ
1

+
? JLI
1

fl X
I

a
1 i

+
X
?

+
*t

a
I
1

X
'1 ool
1

+
X

11
a
1

4>-

Q
*\

<*

lii

3
i

whence

particular

Theorems

for extracting the

roots of equations

may

be deduced.

96. " By this method you may difcover Theorems for approximating to the roots of
furs

242
-pure

^Treatise
powers
%

of

Part II,
of any numlefs

,"

as to find the n root

ber

fuppofe k to b*- the nearefl

root in

integers,
fhall k n

and that k -J-/

n
J

is

the true root, then

2
9

+ nk - f+ n x ?^k*- f*
the

&?r.

= A%

and, afiuming only

two

firft

terms,

f J

= j -i
.

jp
:

or,

more

nearly, taking the three

"

nk

.'

firft

terms,

w-n,

n _Ji-ti"

A-k

n- n

(putting

=.

A k ) =
9

which*

is

a rational

Theorem

for

approximating to /.

You may
of k+fy
viz,

find an irrational

Theorem

for

it 9

by afiuming the

three

firft

terms of the power

72

k*+nk*-*f+nxn

'

-k"*f*z=zA.

For nk n -

f+ n x -Ik *fi

_.

A k* =z m y
you
find

and refolving

this quadratic equation,

2m
n x n

x k* z

/ 2OTH

27?Z ~f~

S
,.

Chap.

9.

ALGEBRA.
/
is

243
when
it

In the application of thefe Theorems,

a near value of
obtained,

then adding

to k, fubftitute the aggregate in place of k in

the formula, and you will,

obtain

a
;

by a new operation, more correct value of the root re-

quired

and, by thus proceeding, you

may

ar-

rive at any degree of exaclnefs.

Thus
k 3
1,

to obtain the cube root of 2, fuppofe

and

f==
nk
n

-\

ml
Til'

= = 0.25.
4
1.25,

In the fecond place,

will

fuppofe

i;=
3

and

be found, by a new operation, equal to

0.009921,
nearly.

and confequently,
the irrational
3

\/ 1

== 1.259921

By

Theorem, the fame

1.

value

is

difcovered for </

CHAP.

244

Treatise

of

Part

II.

CHAP.
Of
the

X.
by which

Method of

Series

you may approximate

of
97.
literal equations.
be only two

to the roots

TF there A in the

letters,

x and a 9
fuppofe a

propofed equation,

equal to unit, and find the root of the numeral

equation that

arifes

from

"the

fubititution,

by

the rules of the

lail

chapter.

Multiply thefe

roots by #, and the products will give the roots

of the propofed equation.

Thus
55
xx

the roots of the equation x

in

i6x-\and 11.
equation
l

o are found,
therefore

84, to be 5 roots
o, will
z

And
The

i6ax + 55a* ==

the

of

the

be 5^ and 11 a.
-f-

roots of the equation x

a x

2a = o
roots of

are found

by enquiring what
l

the numeral equation x

+# =
2
is

are the o,

and fmce

one of thefe

is

1,

it

follows that one of the roots

of the propofed equation

are imaginary.

the other

two

y =
l

98. If the equation to be refolved involves more than two letters, as x -f a % x 2a + ayx
3

o, then the value of x

may

be exhibited

in a feries

having

its

terms compofed of the

their

powers of a and y with

refpective

coef-

ficients 9

Chap.
flcients

10.
,

ALGEBRA.
will

245
I

which
by

"

converge the fooner the

is

in refpecl

multiplied
thofe of

the

of #, if the terms are continually powers of y, and divided by

Ci

a"
is

Or,

will converge the fooner the

greater

in rcfpetJ

of a y if the terms be con-

thiually multiplied by the

by thofe of

y"

Since
1

when y
x

powers of ^ , and divided is very little in

y,

refped of a, the terms

decreafe very quickly.

~, ^ *
in

v*

Sec.

If

y vanifh

refpect

of a, the fecond term

the
firft,

will vanifh in refpedc

of

fince

a.

And

after the

fame manner

vanifhes in refpect of the term

it.

immediately preceding

But when y
then a
refpecl:
0}
-jj,

is

vaftiy great

in

refpecl:

of a %

is

vaftly great in refpect of

and

in

of
in

fo that the terms a.

y
this

*
z

&c_

cafe decreafe

very fwiftly.

In

cither cafe, the feries

fift

converge fwiftly that con-

of fuch terms

*,

and a few of the

firft

terms

will give a near value of the root required.

99. If a feries for x is required from the propofed equation that fhall converge the fooner, the
lefs

is

in

refpecl:

of a
fhall

to find the

firft

term of
nifh
5

this feries,

we

fuppofe

to va-

and extracting the root of the equation

*3

246
x
%

^Treatise
z
%

of

PartIL

+ a x ia
find,

== o, confiding of the remaining

parts of the equation that

do not vanifh with y, by 97, that x == a which is the true value of # when y vanifhes, but is only near its

we

value
little.

when y does not

vanrlh, but only

ftill

is

very

To

get a value

nearer the true value

of

x, fuppofe the difference

of a from the true

value to be p, or that x
ftituting

= a + p.
1Jrf

And

fubx,

+p
T~l

in the given

equation

for

you

will

find,

X*

x + a2a
z
-

3 a *P -1

3 aP

a
1

tzz

2a

4- a p
1

ayx

= y+ apy
zz:

y*
J
a

==

^p + 3^ + ^ 7 _ q
+ apyy^S
<2,

But
little

fince,

by fuppofition, y and
it

/>

are very

in refpecl: of

follows that the terms

are feparately of the

4a zp, a zy, where y and p

leaft

dimenfions, are vaftly great in refpecl of

;

the reft

fo that, in
reft

determining a near value

:

of p) the
z

/\.a

ay

=
a

and from may be neglected o, we find p = So that ~j.

-

x=.a+p
Then

z=z

~y,

nearly.

to find a nearer value of p, and con-

fequently of x 9 fuppofe
ftituting this value for

p=
it

^_y

q,

and fub-

in the laft equation,

you

will find,

fcaP

10.

ALGEBR

A.

247

W ~h ayq = ay f =f
ayp ==
efy
%

4 Z aV - ay + 4a p
%
z

+f

"

=o
I

J
3 11-^*

=
And

tW
+ 4*2

4- Ay*!
E

+
q

+?
is

i=
very
-Jjy,

fince,

by the fuppofition,
which
is

little

in refpect of p,

nearly

there-

fore q will be very

confequently
will

all

of y ; and the terms of the laft equation

little

in refpect

**<&% + A^h
dimenfions

be very

little in

refpect of thefe two, viz.

lead:

where y and q are of

:

\ayq
very
refpecl:

feparately

particularly

the

term
is

is little

in refpecl:

of ^cfq, becaufe y
;

little

of

TV^>%
x

in

refpecl of a

and

it is

little

in

becaufe q

is little

in refpecl

of y. Neglect therefore the other terms, and fuppofing

?

iVzy
;

\o?q

o,

you

will

have

^x?

fo

that x
in the

= -iy + ^ x
fame manner you

?
will

And

by proceeding

100.
for

When
fhall

it

is

required to find a feries

is

x that

converge fooner, the greater y

in

248
in refpect of

Treatise
little

of

Part

II.

any quantity a, you need only fupin

pofe a to be very

refpect of

_y,

and
laft

proceed by the fame reafoning as in the

example on
little.

the

fuppofition of

y being very

a x + ayx / = o
%

Thus, to

find a

value for x in the equation

that fhall converge the

fooner the greater

is

in refpect

of

#,

fuppofe

a
#3

to vanifh,

and the remaining terms

t

will give
is

_- 0j or x y
it

So

that

when y

vaftly

great,

appears that x

nearly.

But to have the value of x more accurately,

put x

=y + p,
x ayx 5=
z

then
y*

+ a^ = =.a*y
1

sy*P d"p

syf +P'

-\-

ay

-f*

ayp

h
,

+
p > and

ay

7
-

ayp

z where the terms %y'p ay become vaftly greater than the reft, y being vaftly greater than a or

Again, by fuppofing p = \a +
laft

confequently

=z

fa

nearly.

you

will

transform the

equation into

a^y

ayq

7,

aq

]than

where the two terms ^qy ay muft be vaftly greater than any of the reft, a being vaftly lefs

Chap.

io.

ALGEBRA.
lefs

249

than y* and q vaftly

fition
;

than a* by the fuppoa*y

fo that

3^*

2
f-

o,

and ^

=
may

nearly.

By proceeding
jy,

in this

manner, you
8tf*

correct the value of

and find that

*3

x
which

= 7 f*H
r

-5^

-,
a.

&c.
is

feries

converges the fooner the greater y

fuppofed to be taken in refpect of

01. In the folution of the

thofe terms were always

determine p,
dimenfions.

q*

r,

quantities p* q* r,

firft Example compared in order to &c. in which y and thofe &c. were feparately of feweft

But in the fecond Example* thofe terms were compared in which a and the quantities

p* q,

r,

&c. were of lead dimenfions

fe-

parately.

And

thefe always are the proper terms

to be compared together, becaufe they

vaftly

become

greater than the reft,

in

the refpective

hypothefes.

In general

*,

to

determine the

firft,

or any,

term

in the feries, fuch terms of the equation are

to be ajfumed together only*

as will be

found
,

to beis,

come vaftly greater than the other terms

that

which give a value of

it

g,

which fubftituted for

above,

in all the terms of the equation fhall raife

the dimenfions

of the other terms

all

or

all

below,

the dimenfions of the aflumed

terms, according as
little,

is

fuppofed to be vaftly

or vaftly great in refpect of a.

Thus


250
Thus
verging

"A

Treatise
firft

of

Part

II,

to determine the
feries

term of a con3

equation x 3

ayx and
For thev

a x + ayx y =
z

exprefting the value of x in the laft

-

o, the terms

are not to be

compared together,
,

would give x

=
3

which fubftituted

for

a:,

the equation becomes

l-

ay

+y y -o,
3

where the
Fewer

firft

term

is

of more dimenfions than

the aflumed terms ayx 9

:

y
firft

3 \

and the fecond of

neither

fo that the

two

terms cannot be nelaft,

glected in Kefpecl of the two

when
a.

y is very Nor are

great nor very

little,
fit

compared with
to be
firft

the terms

x\ ayx 9

compared
term of a
z

together in order to obtain the

feries for #,
3

for the like reafon.

a x with y
l

But x may be compared with

3

x,

as alfo

for that end.

feries

Thefe two give

converges the
gives the
firft
3

the

firft

term of a
lefs

that

fooner the

is

-,

as x

=y

term of a
greater

feries that
is.

converges the fooner the

preceding

article.

The laft feries was given in the The comparing x 3 with - afx
.

gives thefe -two

feries,

VL = ^ a + iy+ f + i6<r + ^ 8a
1
'

W
1280 3
3

&c

The comparing
J,*

'

a*x with.
^4

gives

&c
And

Chap. ib.

ALGEBRA.
\

251
x when

And
y
is

thefe feries give three values of

little

very

the

laft

of which
it

is

itfelf alfo

very

little in

that cafe, as

appears indeed from

the equation, that

lues of x

when y

vanifhes, the three va-

become

->

~~ a

>

an ^ ? becaule
1

when y

vanifhes, the equation

^,

= o,
have

whofe roots are

becomes x
o.

a x
%

102. It appears fufficiently from what

faid, that

wc

when an equation
jy,

is

propofed inis

volving x and
in

and the value of x

required

a converging feries, the difficulty of finding

firft

the

term of the

feries is

reduced to this

M
of

to find

what terms aflumed

in order to deter-

mine a value of x exprefTed

v

in

fome dimenfions
it,

and a

will give fuch a value of

it

as fub-

ftituted for
ail

in the other

terms will

make them
of
lefs

of more dimenfions of

y9

or

all

di-

menfions of j, than thofe aflumed terms."

To determine
angles

this,

draw
it

BA and AC
the

at right

to each other, compleat

parallelo-

gram

ABCD

and divide
In

into equal fquares,

as in the

figure.

thefe

fquares place the

powers of x from
of y from
place that
the line

towards C, and the powers

towards B, and in any other fquare

that
is

power of x
it

directly

below
is

it

in

AC, and
in

that

power of y

that

in a

parallel with

in the line

AB

fo that the inits

dex of x

any fquare

may

exprefs

diftance
in

from the

line

AB, and

the index of

any

fquare

252
fquare

A
may

Treatise
its

of

Part

I&
line

exprefs

diftance

from the

AC.

Of

this

fquare

we

are to obferve,

D %
,

1L7

da V7
y^g-

y7g

-,7&

k M
t/i ycc
6-SL

&&
7/^
tic*
1*3

lk
'foe

ifbc

foe*

&
<?

&
y dc

y oc
(j\zc

t/

4%
4

yS 3 2

foe 7 jc

mJ
2
-4

H& 4?
6\

4 7
*~7 Tx

2.

Ax

3 2

m
s
-J

roc
a-

&' Vcc
ycc7

Vcc

VxHhx^
00

Vcc

s Voc\

x > ^ f

s^fc
\

\E
1.

That the terms

are

not only in geome-

trical progrefllon in the vertical

column AB, or
,

the horizontal

AC, and
for in

their parallels

but

al-

fo in the terms taken in any oblique ftraight

line

whatever

-,

any fuch terms

it is

ma-

Bifeft that the indices

metical progrefllon.
thofe terms will

of y and x will be in arithThe indices of y, becaufe

the line
it,

remove equally from

AC,

or approach equally to

and the indices

of

Chap.

io.

ALGEBRA.

253

of y in any fuch terms are as their diftances from that line AC. The indices of x will alfo be in arithmetical progreffion, becaufe thefe terms equally remove from, or approach to the line

AS.
y>x
z
f

Thus
yto\

for

Example,

in the

terms y 7 ,

y%
x
in-

the indices of

y decreasing by the

common
bers, the

difference 2, while the indices of

creafe in the progreffion of the natural

num.

common

ratio

of the terms

is

y
ec

It

follows,
2.

From
fame
,"

the laft obfervation, that

equal, then
all

if

any
be

two terms be fuppofed

in the

the terms

ftraight line with thefe terms, will

equal

becaufe by fuppofing thefe two terms

equal, the
tio

common
;

ratio is

fuppofed to be a rathis
it

of equality
if

and from

follows, that

"

you

fubftitute every
it

w here
T

for x the value

that arifes for

by fuppofing any two terms

equal, expreiTed in the powers of

fions

y y the dimenbut

of y

in all the

terms that are found in the

be equal

fame ftraight

line will

"

the di-

menfions of y in the terms above that line will be greater than in thofe in that line ;" and

"

the dimenfions of y in the terms below the faid

line will

be
z
\

lefs

than

its

dimenfions in that

line.

5*

Thus, by fuppofing
or x
in

=y

f 2 yx\

we
this

find x z

=y 6

and fjbflituting
z
,

value for x

all

the fquares, the dimenfions of

in the

terms

j, y

x f y*x

yx\ which

are all

found in
the

254
menfions in

d.

Treatise

of

PartIL
but the di-

the fame ftraight line, will be 7,

all

the terms above that line will

be more than
line will

7,

and

in all the terms

7.

below that

be

lefs

than

103.

From

thefe

two obfervations we may

eafily find a

method

for difcovering

what terms

ought to be affumed from an equation in order to give a value for x which fhall make the other
terms
all

of higher^ or

all

of lower dimenfions
:

of y than the aiTumed terms

proper fquares (by the
are to be aflumed as
lie

viz.

"

after all

the terms of the equation are ranged in their

laft article)

fuch terms

in a ftraight line, fo that

lie all

the other terms either

line,

above the ftraight

or

fall all

below

it."

yl

For example, fuppofe the equation propofed is 6 z ax o, then markay 5 x -j- y Ax + a yx*
l

ing with an afterifk the fquares in the

ticle

laft

ar-

which contain the fame dimenfions of x and

terms in the equation, imagine a ruler
revolve about the
as
it

as the

ZE
at

to
7
,

firft

fquare

marked
it

and

moves from
5

towards C,

will firft

meet the term ay

x,

and while the

ruler joins thefe

lie

above

it

poling thefe

two terms, all the other terms from which you infer, that by fupterms equal, you fhall obtain a vait,

lue of x, which fubftituted fqr

will give all

the other terms ot higher dimenfions of

thofe terms
:

y9 than

and hence we conclude that the

equal 3

value of x deduced from fuppofing thefe terms

ehap. 10.
equal, viz.
will

ALGEBRA.

255
feries that

is

the

firft

term of a
lefs

converge the fooner the

is

in refpect

of

a.

If the ruler be

fquare the
will
firft
l

made to revolve about the fame contrary way from D towards C, it

4
jy

meet the term

o,

and by fuppofing

+ y*x =

we

find

y = *,
is.

which gives the

this is the ce-

firft

term of a

feries for #,

that converges the

fooner the greater that

lebrated

And

Rule invented by

Sir Ifaac

Newton

for

this purpofe.

104. This Rule

tions

may

be extended to equa-

having terms that involve powers of x and

or furd indices
in the lines
;

y with /rational
diftances

" by taking
and

from
to

AC
and

AB proand

portional

thefe

fractions

furds,"

thence determining the fituation of the terms

of the propofed equation in the parallelogram

ABCD.
It
is

to be obferved alfo, that

all

when

the line

joining any two terms has

the other terms

firft

on one
<c

fide

of

it,

by them you may find the

term of a converging feries for x,

and thus

various fuch feries can be deduced from the

fame equation."
line joining
it
;
s

As,

in the laft
all

y x and yx* has

and therefore fuppofing

l

Example, the the terms above

ay s x

+ a-yx* Q
is

we $nd x

=L a

and x

= ^-, A

which

the

firft

term

256

Treatise

of

Part IL

term of another converging

other terms above
a*yx*
the

feries for x.

Again,
all

6 the ftraight line joining yx* and x has

the

it,

and therefore, fuppofing

<zx

=zOi

we

find ay

x
is.

%
,

and

#0*7%
are

firft

term of another
the lefs

feries for #,

converging

alfo the fooner

feries

There

two
to

converging the fooner the greater y

6
!

is,

be deduced from fuppofing y 7

~ yx\

or

yV ~ax
fcribe

And, to find all thefe feries, " dea polygon Zakd, having a term of the
.

equation in each of

its

angles,
it,

and including
feries

all

the other terms within

then a

may be

by fuppofing any two terms equal that are placed in any two adjacent angles of the

Found

for #,

polygon."
105.
parallel
If the ruler
itfelf,

ZE

be made to

move
it

to

all

the terms which

will

touch

at

once

will

be of the fame dimenfions of

for they will bear the

fame proportion to one

another as the terms in the line

ZE

themfelves.
firft
it

The

terms which the ruler will touch

will

have fewer dimenfions of y, than thofe

afterwards in the progrefs of
,

touches
if it

its

motion,

but more dimenfions than moves towards D The terms in the they, if it moves towards A.
ftraight line

ZE,
it

ferve to determine the

feries required.

firft

term of the converging

v/ith the

Thefe

terms

touches afterwards ferve to

determine the fucceeding terms of the converging

Chap.
ing

10.
;

ALGEBRA,
all

257

feries

the reft vanifning

is

compared with

thefe,

when y

very

little

and the ruler moves

towards D, or when y is vaftly great towards A. and the ruler moves from

from

106.

The fame Author

gives another

meferies

thod for difcovering the

that
fhali

firft

term of a
lefs

converge the fooner the

the term where

is.

" Suppofe

y
as

is

feparately
it

of

feweft dimenfions to be

By

compare

fuccef-

fively with the other terms,

with Ey mx\ and

*

obferve where
ting

is

found greatejl
be the
firft

and put-

e= ,

Ay n

will

term of a
3'

feries that ihall

converge the fooner the

lefsjy is:'

for in that cafe

By and
1

Ey>x 5 will

be

infinitely

greater than any other terms of the propofed

equation.

Suppofe Fy ex k

is

any other term of

the equation, and, by the fuppofition,

is

(=)

greater than
k,

- e

and confequently, multi/

plying by

you

find nk greater than

I ;

and

nk

+ e greater than
I)
all

4yS
than
fore

then Fy ex k == Fjfifh

now if for x you fubftitute? + % which therefore will

vanifh compared with By! (fince nk

+ eis greater
Thus
there-

when y

is

infinitely

little.

the terms will vanifn compared with -By 1 and Ey mx % which are fuppofed equal ; and
coi.iequently they will give the
feries that will
firft

term of a
lefs

converge the fooner the

is.

107.

2$$

^Treatise

of

Part

II.

107. If you obferve " when

leaft

is

found

of

all,

and fuppofe
fir ft

it

equal to #, then will

feries that will

Ayn be

the

term of a
s

con-

verge the fooner the greater y is." For in that cafe Dy 1 and Ey m x will be infinitely greater than

Ffx*9 becaufe -=
it

(=
is lefs

n) being

lefs

than

-,
K
e

follows that nk
/,

than

e,

and

nk+ e lefs
}

than

and confequently fyxk

(= FA kyni +

vaftly lefs than >/,

when y
if

is

very great.

After the fame manner,

you compare any

term Dy x
l

b
9

with
7

all

the other

where both x and y are found, terms, and obferve where

is

found great eft or Z?^,

then

and fuppofe
term of a
ifc

= ,

may

yfy w

be the

firft

converging

feries.

For fuppofing
if

that Fy ex k

any other term of the equation,

is

(=1 n)

greater than

r,

then

fhall

#&

nh b

greater than /

e9

and

nk-\- e greater than

/+ nb.

But nk+e are the dimenfions of y in Fy'x* when # == yfy", and 1+ nh are the dimenfions of y in Zs/V ; therefore Fy ex k is of more dimenfions of y than iyv, and therefore vanifhes compared to In the it when y is fuppofed infinitely little.
fame manner,
if
r is

lefs

than 7

z>

then
will

Chap. 16.

ALGEBRA.

2^9
when

will Ey m x s be infinitely greater than Fy e x* 9

is infinite.

108.
ries is

term (Jy ) of the fefound by the preceding method, then by

the
firft
tt

When

fuppofing x

^y +
n

/>,

and fubftituting
for

this bi-

nomial and

its

powers

x and

its

powers,

there will arife an equation for determining

the fecond term of the

tion

feries.

This new equaas the

may be

treated in the

fame manner

equation of x 9 and by the Rule of 103, the

terms that are to be compared in order to obtain a near value of p,

may

be difcovered
:

by

means of which terms p may be found

fuppofe
equal
r

which

to

Byn + r,

then

by fuppofing

p =2 By"+ -\- q, the equation may be transformed into one for determining q the third
term of the
of the
feries,

and by proceeding
as

in

the

fame manner you may determine

feries as

many terms

you

pleaie

finding x

= Ay +
n

By+

-f Cy

Zr

+ Dy + *\

&c. where the

di-

menfions of y afcend or defcend according as r is pofitive or negative \ and always " in arithmetical progreffion, that this value of x being
fubftituted for
it

in the

propofed equation, the

its

terms involving y and

powers may

fall

in

with one another, fo that more than one

always involve the fame dimenfion of
jy,

may
the

which

may

mutually deflroy each other and

as
it

make

whole equation vanifh,

ought to do."
Ic

z6<5
It is

A
By" +
r

Treatise
2 >'

of

Part

II.

obvious that as the dimenfions of y in

Ay*

+ Cy" + + By" +
cube,
or

3',

&c. are

in
is

an
r,

arithmetical progreflkm

whofe difference
s
r

the

fquare,

any power
3

^ + r + q + * +

Dj +

of Ay n -f iSc. will confift of

will conftitute-

terms wherein the dimenfions ofy

an arithmetical progreffion having the fame com-

mon
if

difference r

for thefe dimenfions will be

j/z

sn, sn -}- r,

j# -|- 2r,

-\-

$r 9 &c.

Therefore,,
the-

in any term

Zyw you
feries

iubftkute for x
-\9
i

feries

Ay n 4- #j*"+*~i- Cy nJr^
of the

Dy+3 r &c,
Eymx
will

the terms

expreffing

confift of thefe dimenfions

of y, viz.
-j-

m
as

-J-

snt

sn

+ ?% ^ +
e

sn

+ 2r, ^ -f j#
will be e -f
-

jr, &c.

and

by a

like fubftitution in

any other term

;z,

Fy e x kT

the dimenfions of

-f-

nk

+ ?V
fe-

+ + 2r

'

+ ^ + 3*\ & c

The former

ries

of indices mull coincide with the

latter feries,,

that the terms in

which they are found may be compared together, and be found equal with,
oppofite figns lb as to deftroy one another,

and

make

the whole equation vaniflaL

The

firft

feries

confifts

of terms arifing by
to * -j- sn 9 the latter

adding fome multiple of r

by adding fome multiple of

thefe

r toe

+ nk

and that

may

coincide,

fome multiple of r added fome other multiple From which it appears,

to sn

to

m 4- sn

muft be equal
e

of r added to

+ nk.

that the difference of

m+

and e-\-nk

is

always

a mui-

Chap.

10.

ALGEBRA.
and confequer.tr/ that r

261
is

a multiple of r;

divifor of the difference of dimenfions

of y in

the terms
It follows

Eymx* and Fyex*, fuppofing x =: Ay". therefore " that r is a common divifor

of the differences of the dimenfions of y in the terms of the equation, when you have fubftituted

Ay n

for x in all the terms."

afTumed equal to the great eft

And if r be common divifor

feries
r
,

(excepting fome cafes afterward to be mentioned)

you

will

have the true form of a

the

for x.

And now
jy-f-3>',

dimenfions y",

y+

n ~^" 2r
7

&c. being known, there remains only,

to

by

calculation,

determine
Z),

the
in

general coto find

~|-

efficients

C,

&c.

order

the feries

&c.

= x.

Ay n -c By n + r ~\- Cy-V n r _L Dy nJr

109. This

leads us

to

Sir Ifaac
-,

Newtotfs
confifls

fecond general "method of

in

feries

which

afTuming a

feries

with undetermined coeffici-

ents expreiling x,

as

Ay

i:

-f By"

+r

+ Cy

+
as

2~

-f-

&c. where

B, C, &c.

are fuppofed

yet

unknown, but n and

every where for
x,

r are difcovered
;

by what we
this

have already demo nil rated

and fubftituting
fuppofe, in

you
arifes,

mud
the

the the
jy

new equation

that

fum of

all

terms that involve the fame dimenficn of

vanifh, by which means
lar equations,

to

you

will obtain particu-

the

firji

of which will give A*

being

the fecond B? the third C, &c. and thefe values

S 2

262

^Treatise
feries for

gJ

PartlL
feries

being fubftituted in the affumcd"

for

A,

By C, &c. the
far as

x will be obtained as

you

pleafe.

Let us apply,

for

example,

this

z >ia y -\-ayx a x y> the equation (of 98) x z Suppofe it is required to find a feries o,

+
is

method to

converging the fooner the

lefs

is

its

firft

term (by 99 or 102)

that n

found to be .a 9 fo
1

o.

Subflitute a for x in the equation,

and the terms become /r

+a

la

-\-

ay

+
x

jy%
;

and the differences of the indices

are o, 1, 2, 3
is

whofe greater!
r

common

meafure

1,

fo

that

= + Dy

1.
l

AfTume
&c. and

therefore x
fubilitute

= A+
feries

By

Cy z
in

this

for

the equation.

Then
l

#*=

A*-\-3A By-\-3/fBy-{-

By

-f &

+3^Cy*+ 3 ^iy+^.

+ 6ABCy*+&c.
-}-<j
a
tf

= <r^+
J

By 4- *Cj z

-f-

;rM -f &V.

2# = 2a J =
3
3

-ix/.
+ a*x + <gw 2# y* = o,
5

Now
mufl
ficient

fince #*

it

follows that the

vanifh.

fum of

thcfe feries involving y

if

But that cannot be

the coef-

of every particular term does not vanifh.

is

For every term where y

finitely greater

infinitely little,

is

in-

than the following terms, fo that

if

Chap. 10.
if

ALGEBRA.
itfelf,

263
the ad-

every term does not vanifh of

or
fubtraction

dition

of the

following
it,

terms

which

are infinitely lefs than

or of the pre-

ceding terms which are

not deftrcy
vanifh.
It
is
it
;

infinitely greater,

can-

and therefore the whole cannot

appears therefore that A>

a*-

2a> == o,

an equation for determining A, and

gives

A a.
you muft fuppofe
the coefficients affecting
or,

In order to determine B,
the
viz.

fum of

to vanifh,

SA-B

+ tfB + aAxy O)

fince

A-=za y

4a-By

a"-y

= o,
for

and

B ==

To
pofe

determine C, in the fame manner fupl

z a Cy

%ABy + sA Cy* +

+ aBy* = o,

or,

fubftituting

A
z

and
z

their

values already

o, and confe *- + 4a Cy And, by proceeding in the quently C = - fame manner, D = ^-r, fo that x = a iy
found,
.

+7
99-

1
-\

'"t.

l!

&c.

as

we found

before in

ries

no. By this method you may transfer fefrom one undetermined quantity to another,
for the reverfwn
z

and obtain "Theorems

Suppofe that x

of
4

feries.

= ay + h +
It

cy

+ ^ + &ta
feries

and

it

is

required to exprefs

y by a
is

con-

fiding of the powers of x.

obvious that

S 3

when

264
when x
feries,
is

^Treatise
very
little,

of
very
firft

Part II.
little,

is

alfo

and

that in order to determine the

term of the
there-

you need only aflume *

;

ay. And
By

fore

y^-1
you
i,

fo that n

i.

fubftituting

for j,
will

find the dimenfions of x in the terms

2,
3,

be

4, C5V.

fo

that r
=.

s*=
-\-

alio.
-j-

You may.
Cx
l

therefore aiTume

Ax

Bx z

ZXv 4

ffr.

And

by the fubftitution of

this value

of y you will find

#i?,
a
3

= aAx + + aC% &t. by* = kA x + 2bABx> + fcfo

ay
-\z z

ry 5

cAx

+ sV.
'0c.

h?c.

But the

firft

term being already found to be

a
it

you have
follows

A =z a
B

and
-

fince

aB

+ bA =
z

o,

that

=
n

-^.

After the fame

.

manner you
x
b
^

will find
x

C ==
ac

.
.

Whence

ib -

111. Suppofe again you have ax

r#
3

+ b>r
for

-j-

J,v 4 -|-

&V.

= gy +
firft

iry

zj

+ ^^

&c. to
by

find # in terms of y.

You

will eafiiy fee,

feries

103,

that the
that

term of the

is,
a

i,

1.

Therefore

affume

Ay + By*+

Cy\ &c. and by

fubftituting
this

Cliap. 10.
this

ALGEBRA.
all

265
the terms to

value for x and bringing

fide,

one

you

will

have
%
1

ax == a Ay

ix

cx> z=

+ aBy + aCy + bAy + ibABy^ + cAy- +

be,

&c* &c.

&c*

fc

gy = y

y=
C5f.

............
fee,
firft,

iy ?

fc?A

From whence we

that
2

aAg^

and and

^L a

2*.

That

<?5

= o,

Bz=.-

-.

r .i;\Y?xaC+2bAB + cAi = o,
C
=.

and therefore
three
firft

a
feries

And
1

thus the

terms of the
*.

Ay + By

+ Cy\ &c+

are

known

112. Before

we conclude
method

it

remains to clear
Ci

difficulty in this
late

that has embarraifed

fome

ingenious writers, concerning

the

value of r to be aiTumed

when two
;*'

or more of
for ex-

the values of the

firft

term of a
equal

feries

preffing x are found

a correction of

the preceding Rule being neceffary in that cafe.

And

the author of that correction having only

it

collected

from experience,
De

and given

it

us

* See Mr.

Moi-ure in PhiL Trcwf. 240,

S 4

with-

266

A
it

Treatise
it
is

of

Part

II.

without proof,
monftrate
It is to
feries

the

more

neceffary to de-

here.

be obferved then, that in order that the

--}-

Ay"

By*

+r

-j-

Cy n +

2r

4-

T + y + & c
m

may
it
is
l

exprefs #,

it is

not only neceflary that

when
-\~ ns^

fubftituted for x in the propofed equation

-f-

Dy

Ejpx*

-j-

Fy

,\

o,

the indices

m-\~ns-\-i\ m-^ns-^-zr, &c. fhould

the indices e-i-nk,
in order that

fall

in with

e\nk~\-r, e-^r.k-\-2r^ &c. the terms may be compared toge&c.

neceflary, that in

ther to determine the coefficients A, B> C,

but

it

is

alfo

the

particular

equations for determining any of thofe coefficients,

as

for

example* thofe terms that

in-

volve

fhould not ceftroy each other.

Thus
o,

the equation

^A^B ^A'B
becaufe
itfelf

aA. o
Z

determine
thus

i?,

^A% B

3A B
from

can never

and
;

exterminates

out of the equation

befides the contradiction arifing

A=

o,

when

A perhaps

has been determined already to

real

be equal to fome
In order to

quantity.
to evite this abfur-

know how

dity, let us fuppofe that the firft order of terms

the propofed equation are, as before, Ey mxs, &c. and if Ay* is found to be the
in

Dy !9
firft

term

of a feries for
firft

>:,

then the dimenfions of

in the

order of terms, arifing by fubfrifor x. will be

arifing

tuting in
the

them Ay*

-f-

and

dimenfions

of y

by

fubftituting

Ar

Chap.

io.
r

ALGEBRA.
4- Cy +
2r
,

267
be

Af 4- By +
m + ns + r,
is

&c. for x

will

m -\- ns,

4 j 4 2r,

&c.

Suppofe that Fy ex k
of y arifing from
r

the next order of terms, and, by the fame the dimenfions

ftibftitution,
it

will

be

(becaufe F) ex*

Fy x Ay+ By+ + Cy+ + &c] k

2r

s= FA y + k
k

kFBA*-y + * +
nk

',

&c.) *
it is

nk>

-\-

nk-\- r,
*

e -J-

2r,

&c.

Now

plain

that

+ #

muft coincide with fome of the

m-\- ns 4 2r, &c. them may be compared
as

dimenfions

m 4 ns, m + ns-\-r,
And
therefore,

that the terms involving

together.

we obferved
e

in

108, r
in

mu ft

be the difference of

4 nk

and
In

4- J, or

fome

divifor of that difference.

general, r

muft be affumed fuch a divifor of that

difference as

may

allow not only e

4 nk
m
as

to co-

incide
;

with

fome one of the

ns
-\-

feries

4
-\-

#*

m 4 ^ 4 ft m +
all

2r,

&c. but

may make
nk
feries

the indices of the other orders befides e

to
if

likewife

coincide

with one of that

that

is,

Gyfx h
fo

is

another term in the equaferies

tion, r

muft be
r,
it

affumed that the

4" 2r>

/ + ,
by fubfor x,
feries

f -\-nh~\may

/4- 'ih

&c

a riQng
,

ftituting in

Ay n -L By + r -. Cy n + zr &c.

coincide
tot
ni

Lmewhere
7

with

the
2r,

firft

rt

4^4
m

&c.

And

therefore

fumed

fo

we faid, in 108, " that r muft be afas to be equal to fome common divifor
-\- ns,

of the differences of the indices

e -\-

nky

f -\- nb 9 which

arife

in

the propofed

equation

by

268

Treatise
x the

of

Part

II.

by fubftituting

in it for

firft

term already

known Ay n " For by

afliiming r equal to a

com-

mon

divifor of thefe differences, the three feries

9

m + ns m+
3

+ 2f m+ ns+ ^r, &c. + nk + 2r e + nk+^r^&cc. *-f &, e + nk + ^ /+ + /+ ^ + 2r, /+ #& + 3^, &c. /+
ns -f r, 0* -f #*
r, e
,
9

to&

ta

will coincide with

tiples
all

one another,

fince

of r added to
it

+ *m will
m+
it

fome mulgive g + and

and fome

that follow

in the fecond feries,

multiples of r added to

n* will alfo give

f+nh
hinder

and

all

that follow

in the third feries.

It is alfo obvious, that, if

it,

no particular reafon

r ought to be afiumed equal to the

great eft

common
if

meafure of thefe differences.

the indices

For example,

+ ns,

nk y

f+nh, happen

to be in arithmetical progreffion,

then r ought to be afiumed equal to the com-

mon
the

difference of the terms,

feries will

and the

firft

of

the fecond
firft,

coincide with the fecond of

and the

firft

of the third
and

feries will
feries,

co-

incide with the fecond of the fecond

and

with the third of the

firft,

fo on.

113. Thefe things being well underftood,

we
firft

are next to obferve that after

you have fub-

ftituted

Ay*

By n + r

Cy-f"*9 &c. for x in the

order of terms in the equation, the terms

that involve

m 4for

ns dimenfions of

Be another;

x Ay

will deftroy

mu&

be a divifor of
the

Chap. 10.
the aggregate

ALGEBRA.
let

269

of thefe terms, fince they give x

Ay n as one value of x:

Ay

x P reprefent
its

that aggregate, and, fubflituting for x

value

Ay n -\- By nJr rJr

2

Cy"^r ir ,

&c. that aggregate be-

comes Ay n + By + r +Cy Rjt zr , &c.

Ay"

x P

By+

+ C>*+

dirnenfion in

&c x P- Now the loweft x Ay x P was fuppofed to be

2r
,
tt

m + ns,
fion in

whence the dirnenfion of P, in the fame terms, will be m ns n, and the loweft dirnen-

By+ + Cy r -{-&c. x P will btn + r m + m n~m~\-ns + r. Suppofe again that +

r

n ~J

zr

two values of x 9 determined from the

of terms, are equal, and
then x

firft

order

Ay"\
firft

will

be

a divifor of that aggregate of the terms. Suppofe that aggregate

order of
z

now x Ay\ x P, which by fubftitution of Ay*+ By n + r + Cy n + Zr &c.

,

for
in

will

become

2ty

-t- '

-L.

Cy+ ir -{-&c.Y

P*

which the loweft term

will

now be of m
l

dimenfions, fince in x

Ay\
+

+ ns
and

x P

the loweft
;

term

is

fuppofed of

m
in.

ns dimenfions

confequently, in thefe terms, the dirnenfion of

itfelf is

-\-ns

In general^

if

the

number of values of x fup-

pofed equal to Ay n be p, then muft x

Aytf be
firft

a divifor of the aggregate of the terms of the

order.

And

that aggregate being expref&d

by

Ay\ p

Pi in the loweft terms, the dimenfions


A
y
in

zjo
fions of

Treatise
will

of
9

Part

II.

be

+ ns pn
as
n

that in

Jv\? they

may be m +
~i

pofe.
its

Subftitute in

^ Ay x
]

we always
for

fupn

X P
fcfo

Ay
in

value By n r r
n

Qy+ 2r
2r

and

the

refuk By

-\-

Cy+

-j-

&c\ p x

the

lowed

dimeniions of y will be pnA^pr-^-m-^ns

p
firft

114.

From what

has been faid

we conclude,
the
firft

that when you have fubftituted for x in the

order of terms of the equation propofed

feries

Ay n

-\-

By n

+
is

-\- Cy 4- *; if,

&c.

the

term of which Ay"

x whofe number
m-\-ns^-2r, &c.
will
firft

is

known, and

the values of

are found equal, then the

terms arifing that involve towns',

till

m-\-ns\r y

you come to m~\-ns~\-pr y deftroy each other and vanifti ; fo that the term with which the terms of the fecond orcan be compared muft be that which
;

der

e+nk
,

involves
.

m + ns + pr
1

and therefore fuppofing

e?^nk-=i0-\-ns~T-pri or r

= e-\-nk

m
ns

,,

"the

higheft value you can give r muft be the difference

of

+ nk

equal values of the

If this value of r

and m-\-ns divided by p the number of firft term of the feries."

is

common

meafure of
is it

all

the differences of the indices, then

a juft

value of r

but

if it is

not, fuch a value of r

this

muft be affumed,
the differences
:

as

may meafure
is,
**

and

all

that

fuch a value as

may
be

Chap.

io.

ALGEBRA.
common
p
meafure of the
leaf!:

271
dif-

be the greateft

ference divided by

(viz.
all

and

of

the.

common
&c.

meafure of

the differences."

For thus
-f- 2r,

the indices m-\- ns 9 m-\-ns-\~r 9 m-\-ns

will coincide

with e-\-nk 9 e-\~nk\~r9

e-\-nk-\-ir, ccc. and with f-\- nb 9

f-Unb -\-

f~\-nh
to be

-\-

2r,

&c. and you

the
x.

fhall

always have terms

compared together
affumed for

iufficient to

determine

, C, D, &c.
ferics

general

coefficients

of the

Ay

115,
n

To

all

this it

may be

added, that

if

be a divifor of the aggregate of the

terms of the fecond order Fy ex* 9 &c. then, by By n + r Cy n + 2r fubftituting for x the feries Jyn

-f &V. there vanifh not only

feries

as

many terms of the

9
9

involving
as there

m + ns m + ns + r m + ns+2r
9

&c.

are equal values of the

firft

term

Jy n

but the terms involving e-\-nk dimenfions

alfo
;

of y vanifh
fo
that,

and therefore

it is

then only ne-

ceffary that e-\-nk-\-r coincide with

in

m + ns+pr
only
n

that

cafe,
.

you
if a

need

take

And

dy Y
fecond
2r
)

be a
order

divifor of the aggregate of the Do O

of terms, then the terms

(after fubftituting for

n

x the
volve
ni.fti

feries
e

dy n -\-By
9

+ nk

+ Cy + &c. which ine + nk + r + nk + ir &c. will va^

T,

-r r
9

to the term -j- -["?

ixr;

fo that,

fup-

272
fuppofing
e

^Treatise
1

of

Part IL

have re+nk m

+ nk+p x r = m + ns+p-,
ns9 that
is,

you

to the
r

leaft dif,

ference of the indices m-\-ns 9

+ nk, + nh
may
feries

&c,

provided that difference be a meafure of the

other differences
;

although there
firft

be as
equal

many

values of the

term of the
Or,
as

as there are units in p.

if that

does not
equal
diffe-

happen, r muft be taken,

to the greater!:
rences.

formerly,

common

meafure of the

the

third

Jy f x P, by x Ay
r

116. Suppofe that the orders of terms of equation can be expreffed the firft by
the fecond
n \*

by x
xk

Ay
on

n q
\

>,

the
that

L, &c. and fuppofe

Fy
e

Eymx s

is

one of the

firft,

one of the fecond,

fo
:

Cyhh

one of the

third,

and

then

it

is

plain that, fubftituting for x the feries Ay* 4By nJr r -\- Cy*+ 2r &c. the loweft term that will
,

remain in the

firft

will be

m + ns+pr dimenfions

of j, the loweft term that will remain in the fecond will be of e nk qr 9 and the loweft term

+ +

remaining

in the third

of /-f nh-\-lr dimenlions

as

of y.

For by the fame reafoning

we

ufed, in

113, to demonftrate that, in the

terms x

firft

order of

Ay f x P,
n

the loweft dimenlions of

y are

4.

ns+pr, we

mall find that, in the fub-

fequent orders, the loweft dimenfions of y in the

the terms x-Ay*\ q x gjzz ~&y+r

+ Cy+*

8cc.\* x

muft

Chap.

ALGEBRA. = muft be + nk qn + and of the other terms x Af\

io.
e

273

-{-

on

-f-

^r

-f-

jr>

fo

x
/r.

the

lowed dimenfions
nifh being

mull: be

/+

nh 4.

The

indices therefore of the terms that

do not va-

* *

+ ns -\-pr
e

* * * *

******
if

-f nk

+ qr, f + nb + lr

r be taken equal to

^t

'

then will

w + j+/>r

and

+ nk-\-qr
a
1

coincide: and if at

the fame time r be a divifor of/-j-#&

m nsy
ns
~

and be found
,

in it
t-

number of times
1

greater
%

tnan

/,

or

if r

be

lets

f Ar nh m than/
1

then r will be rightly afTumed.

all

In general*
f-\-nh J
,

"
;

take
,

w M the quotients
T
.

*-4-/

573

-~

m ns

and

either the leafi of thefe, or a

denominator, exceeding

number whole
by an integer,

m ns,
gers.

meafures

it

and

all

the differences

/ + nh
you

gives r,"
if

fuppofing

q,

and /inteare to

But

p,

q,
it

and

are fractions,

"

take r fo that

De eaual to

p
anc*

q-\~

^ K

'

ai

m *
'

anc*

**

^ At

ml ma y k e

integers."

Suppofe, for example,

m-\-m = ~

274

^Treatise

of

PartIL

# = ~i e4-nk
/

f-Unh

^-,

and

:=

then putting
I 3

I+Z
M=z-r~h \?K'
and
For
1 1
9

/>

/+^f
is

"" 2

whence

it

eafily feen that 5

are the leaft integers that can be afliimed

K and' M.

And

that r

^ =
i-f-iT

-i

: '

and and
firft

therefore

^+^i+^r= e+nk-\-qr=Q 12 12
,

f+nh+lr.
feries

That

is,

the terms of the

whofe dimenfions are m-\-ns

+ p-\-Kxi\
firft

m + ns +p + Mx
of the fecond
* See on
Mowrei.

r fall in with the

terms

and

third feries refpe&ively *.

this

fubjeft, Col/on.

Epifl. in
Stirling

Animadv. D.
Lin.
iij

Taylor Meth.

Incr.

Ord

iGravefande Append. Elem. Algebrae.

Quadrature of Curves.

Stewart on the

CHAP.

chap. ii.

Algebra.

275

CHAP.
Of
ber

XL
numin

the Rules for finding the

of impoflible Roots

an

equation.
117.

flpHE

number of

impojftbk roots in

an equation may, for molt part, JL be found by this

RULE.
" Write down
2, 3, 4, 5,
exprejfes

feries

of fractions whofe denomi-

nators are the numbers in this progreffion i,

&c. continued

to the

number which Diorder

the dimenfion of the

equation.

vide every fraclion in the feries by that which

precedes
it,

and place the

quotients

in

over the middle terms of the equation.

And

if the fquare of any term multiplied into the

fraclion

that Jlands over

it

gives a produtl

greater than the reclangle of the two adjacent

terms, write under the term the fign
-f-,

but if

that pro duel

write

is

not greater than the reelangle, the figns under the

there will be as
there
,

and

extreme

terms being

many imaof the

-{-.

ginary
figns

roots

as
to

are

changes

from

and from

to

Thus,

2y6
^p x
the
z

^Treatise

of

Partll.
-f-

Thus, the given equation being x 3

px z

+
of

z=z

o, I divide
,

the fecond
firft,

fraction

feries --,

123

by the

and the third

by the fecond, and place the quotients

arid

over the middle terms in this manner

T
xz
-f-

px l

+
Then
that

2p x

q=
+

o.

becaufe the fquare of the

fecond

term
it,

multiplied into the fraction that ftands over

is,

xp
the

x+

is

lefsthan ^p z x 4 the rectan-

gle under

firft

and third terms,

under the fecond term the fign

I
:

place

but as

~X
3

gp*x z

(=

$p*x 2 ) the fquare of the third

its

term multiplied into

nothing,

fraction

is

greater than
greater

fqx
terms,

and confequently
write under

much

than

the negative product of the adjoining

I

the third

term the fign

Jr.
firft

write
laft

likewife under x 3
;

and
in

q the

and

terms

and rinding

the figns

thus marked two changes, one from

and another from

tion has

+
a

to

to -p, I conclude the equaroots.

two impoflible

In like manner the equation x l

4# + 4.x

has two inipolUbie roots

-,

T
3

T
-J-

__ 4#*

aj

and

Chap.

ii.

and the

ALGEBRA. equation #** 6x %%

%
\

'zyj
2

=o

the

fame number

T
#4*

T
#*

T
2^
2

++ +
For
by dividing them
5

= 0.
,

+
,
~

the feries of fractions

as the

~
1

yields,

Rule

directs, the fracti-

ons ~,

~ to be placed over the terms.

,

Then

the fquare of the fecond term, which

is

nothing,
ftill

multiplied by the fraction over

nothing,

it
6

and yet greater than

6#
I

being

the negative

product of the adjacent terms,

(*) the term that
is

write under
-j-,

wanting, the fign

and

proceeding

as

in

the former examples, I con-

clude, from the two changes that happen in the

feries

++

that the equation has

two

of

its

roots impoflible.

The fame way we

roots in the equation

difcover

two impofftble

4* + +
#5

4 ~{-

zx* 5# 4=0. - + + +
4* 1
are

When
fign

two or more terms

firft

wanting

in the

equation, under the

of fuch terms place the

under the fecond

fo

+, under

the third

and

on alternately, only when the two

left

terms to the right and

of the deficient terms

2

have

278
have contrary
lign

A Treatise
figns,

of

PartIL

you

are always to write the

+ under
x5

the laft deficient term.

As

in the equations

and

+ ax* * * * + = o + + + + x + ax* * * * a = o + + ~+ + +
5 s

the the

fir ft

of which has four impoflible roots, and

two.
5
*9~

other

Thus

2x + 3# 2X + * * * ^ o + - + - +-+ +
7

3353 T
"9"

likewife

the

equation

has^
roots

impcflible roots.

Hence
lie

too hid

we may difcover if the imaginary among the affirmative, or among

For the figns of the terms which fland over the figns below that change and from to to +, fliew, by the numthe negative roots.

ber of their variations,

lible roots are to

how many of the impof*,

be reckoned affirmative

and

that there are as

many

negative imaginary roots

as there are repetitions

of the fame fign.

As

in the equation

xs

4*
+

+
_|

4* ?

zx x 4 = o
%

'

+
-j

the figns (

2x

of the terms

4.x*
f-

+ 4^
point-

which Rand over the figns

ing

Chap. ii.

ALGEBRA.
in the

279

ing out two affirmative roots *, we infer that

two impoffible roots lie among the and the three changes of the figns
tion

affirmative

(+

equa-

giving three affirmative

one

two negative, the five roots will be two negative, and two imaIf the equation had been ginary affirmatives.
roots and
real affirmative,

variation
fign
,

ax* 4* 2x 5* 4 = + + + + Hand over terms 4* 4** of fhew, by one imaginary be and terms ix $x reckoned
xs
3

0,

the

that

the

firft

-}-*

the repetition
root
is

the

that

to

negative,

the

that fland over

for

the

laft

variation

(-,

give,

the

fame
,

reafon, another negative

impof-

fible root
(-j

fo

that the
)

figns

of the equation

giving one affirmative root,

we
two

conclude that of the four negative roots,

are imaginary.
unlefs,

This always holds good

times

which fome-

may happen,

there are

more impoffible

roots in the equation than are difcoverable by

the Rule."

This Rule hath been inveftigaied by fever al eminent Mathematicians in various

ways ; and

others',

fimilar
* See

to

it7

invented and -publifhed t.

But the

19.

f See Stirling's Lineae iij. Ord. Iranf. N 394, 404, 408.

Nenton. p, 59,

Phil.

original

iSo
original Rule

Treatise

of

Part

II.

being,

on account of

its fimplicity

and

eafy application, if net preferable to all others*

it is fufficient

at leafi the fittejl for this place,

diretl the

to

Reader where he may find the fubjtel

,

mere fully treated

i

and

to
it

add the demonftration

towards the end of his

w Author
>er

kaj g ven of
only on

to

Mr, Fcikes,

Phil. Tranf.

N 408,

as

it

mds

what has

been demonftrated in

Chap.
tions,

5. concerning the limits of the roots of equa-

118. Let ax 2

+ px + q =
%

o be any adfecTed

quadratic equation
roots will be
is

and, by
%

%%%, Part

I.

its

x T i ^p +
j>

4^- whence it

plain that, the iign of q in the given equation

being -f , the roots v/ill be impoflible as oft as 4aq is greater than p~, or ~p z leis than a x
119. It was fliewn, in general
that
n

the

Bx - Z
n2
is

C#
n

roots
8

of
3

the

equation

xn

&c.

=
-=z

(45 50) Ax* +

J

o, are the limits of

-~ J

the-

re ots of the equation nx n

n
+

~ X Ax n 2

Bx ~i &c.
it

o, or of any equation that

its

deduced from

by multiplying

terms by any

arithmetical progrejjion

1+

d, l

2d,

IT 3d,

&c.

and

converfely the roots of this

new

equation will

be the limits of the roots of the propofed equation

x*

4**-!

+ B$"~z &c. =: o.
]

And
limits

that if a

iy roots

of the equation of the

are impojhbk,

there mufi be

fome

roots

of

the ^propofed equation impojjible.

I 120.

Chap. ii.
120.

ALGEBRA. Bx C == o Let x Ax
3

281
be a
limits

cubic

2x z

2^-f5 =
laft

equation,

arid
o.

-f the

equation

of

If the tv/o roots of this

laft are

imaginary, there are two imaginary roots

of the given equation x*

Ax
z

U-

Bx

C
By

o,

by the
this

Art.

But, by the preceding Art.

happens

as oft as

^A

is

lefs

than

and,

in that cafe, the given equation has

two imagi-

nary roots.

Again,
tion

multiplying the

terms of the equa-

we another Ax 2Bx + 3C = o whofe

2,

by the terms of the progreffion, o,

3,

t,

get

equation of the limits

two
than

roots,

and

confequently two roots of the given equation,

are imaginary

when ~B Z
o,

is lefs

Hence likewife Bx z Cx + D =

the biquadratic
v/iil

x* Ax +
3

AX

C.

have two imaginary

roots, if

+
4#

iBx

of the equation Ax
be imaginary.
?

C =l o be imaginary 2Bx + $'Cx ^D = o

;
1

two

roots of the equation

%Ax z 4# J or if two roots

3C

But two

nary,
3 Ax
-)-

z C = o muft be imagi3 Ax + iBx when two roots of the quadratic 6x z

roots of the equation

+ B = o,

or of the quadratic

3^* 4##

n o,

are imaginary, becaufe the roots of

are

thefe quadratic equations

the limits of the

roots of that cubic

and for the fame reafon

two
opx

4D = o

roots of the cubic equation Ax?

iBx* +
o,

muft be imaginary, when the

roots of the quadratic

%Ax z

^Bx + 3C =

or

2S2

A
3

Treatise

of

Part II,

z %Cx 6D o are or of the quadratic Bx Therefore two roots of the biquaimpofiible.

dratic x*

Ax + Bx

imaginary when the roots

three quadratic equations

$Ax z

aBx + 3C == o
;

+ = Cx + D o muft be of any one of 6x +B=

z

thefe

o,Ax

o,

Bx z
is,

3C* + 6JD == o
when

become imaginary
than 5, **
i2i
lefs

that

4^

is

lefs

than

AC,

or -JO
in

lefs

than BD.,

By proceeding

the fame manner,

Ax" 1 you may deduce from any equation x n -}- Bx"~ 2 o, as many quadratic Cx n i &c.

equations as there are terms excepting the

firft

and
if

laft,

whole roots muft be

all real

quantities,

the propofed equation has no imaginary roots.

The
x
n

Ax* +
l

Quadratic deduced from the three

firft

terms
this
%

Bx 2
n
1

will manifeftly 3

have

form,

11

xn X
?z

X x

X x X n 3 X n 4 &c. x -^ X 4 X 72 5 &c. X 5 =

2 X n
n

&c.

+
o,
as

continuing the factors in each

till

you have

many
n

as there are units in

all

2.

Then

divid-

ing the equation by

4,

the factors
in

n 2, 11r%*
n-

&c. which are found

will

each coefficient,

the equation

become n
1

x % -~
will

x 2 Ax -f

x X B = o,
X a4"'*
or

whofe roots
i

beimaginary, by 118, when

exceeds n
Ptrrl

nx n

x 2 x 4-3

fo

il*

when B exceeds
equation muft

z
:

that the propofed

have

Chap. ii.
have fome
n

ALGEBRA.
imaginary
roots

283

when

exceeds

z
.

The
T

quadratic equation deduced in the

firft

equation Ax"- ?.Bx + %Gx 3 &c. = o, form, n Xn 2x &c. have X Ax n2 x n x n 4&c. x iBx which n X n 4 X # 5 &c. x 3C o
2

fame manner from the three

terms of the

will

this

-{*

?.

dividing by the- factors

js

common
n

to

ali
7 -

the terms,
2

reduced to n

x Ax

\Bx

6C

=
ii
'

when
fore

1 3 in that cafe

2 X

c, whofe roots 2 l

mull be imaginary

XB

is lefs

than yfC; and there-

fome roots of the propofed

let

equation muft be imaginary.

Fx ~ r be any three terms xn Ax n ~ + Bx n ~* &c. =s o,

n
l

122. In general,
J

Dx n ~r +

E&-*
n

-J-

of the equation,
that immediately

follow one another

multiply the terms of this

equation

firll

by the progreffion

&c. then by the progreffion n 1, n 2, n 3, &c. then by n 2, n ^4, &c. till you 5, n have multiplied by as many progreflions as there

n
,

1,

2,

are units in n

r
2, 3,

then multiply the terms

as often

of the equation that

greiTion o,

arifes,

by the prounits in

1,

&c.

as

there are

1,

and you

will at length arrive at a

qua-

dratic of this

form

-{1

X nr X r x n - r - 2 &c. 2xr-jXr-4 &c. X Dx z X r

1

A
r

Treatise
i

of
2

Part

IL

;/-rx n X nr x n-r-z<kc. X X X X &c. x ~\-n X n X x#-r~4 &c. xr+iX^Xr-iX &c. x F=o: n n by and
r
r
1

Est

?z

dividing

the factors

&c. and r
each

1,

1,

25

2,

&c. which
2

are

found in

coefficient, this equation will

be reduced to

r -L.i

X ^^ r x 2X xDx
1

n rX 2-X t
whofe roots muft
.

x 2 Ex

+2x
than

xr

-j-

X ^F= o,
8,

be imaginary, by
is lefs

1 1

when __
it is

x-

X E

DF.
n

From which
tf^3

manifeft, that

if

you divide each term of

n

this feries

of fractions

n
1'

- r-j-i
r

by

2 4 3 that which precedes

n "~ 1

r+i,

it,

and place the quo-

tients

Ax

above the terms of the equation x n Bx* % ~~ C^~3 &c. o, beginning with

the fecond
tiplied

then

if

the fquare of any term

it

multhan

by the fraction over

be found

lefs

the product of the adjacent terms,

roots of that equation
tities.

fome of the

mud

be imaginary quan-

123.

An equation may
though

have impofilble roots

:

although none are difcovered by the Rule

caufe,
*'

be-

real roots in the given equation

always give real roots in the equation of limits \ yet it does not follow, conve?fely, that when the
roots of the equation

of limits are

real,

thofe

of

Chap.

n.

ALGEBRA.
it is

285

of the equation from which be fuch likewife.

produced muft

Thus

the cubic

^ ~~

2m

qy

lxx z+ m% +2qm>XX qXt +n = o + n J

I

has two of
s/

its

roots imaginary,

m
q
:

m n
the
z

-f-

V n
x
z

the third being

equation

of limits $x
>

+
40*

and yet

in
-{-

+HX
im

m +

iqm

+^=

if ^ jl

exceeds 3^,

the

roots of the equation of limits will be *eal.

if the

Or

other equation
J

2X
is

2qm
it
~

+ q x x* + nxx-\-3qX?n + n = o
of limits
1

2, 3
1,

found by multiplying by the progredion o,

;

will

have

its

roots real as oft

:

m And
as

~f-

iqm-\-n\

exceeds

the like

may

3q X m --f-n> be fhewn of higher equa-

2m + q X

tions.

124.

The

reafon

why

this

Rule, and perthe compari-

haps every other that depends on

fon of the fquare of a term with the rectangles

of the terms on either

times
fail

fide

of

it,

muft fome-

to difcover the impoffible roots,

this consideration
;

may

appear likewife from

that the

number of fuch comparifons being always lefs by unit than the number of the quantities q 9
m,
#,

&c.

in the general equation

they cannot

include and fix the relations cf thefe quantities,'

on

a86
on which the

Treatise
of greater or

of
lejfer
\

Partll.
inequality

ratio

of the fquares and rectangles depends

no more

than equations fewer in number than the quantities

fought can furnifh a determinate folution

of a problem.

CHAP.
of
Sir

XII.

Containing a general demonstration

Ifaac

Newton 's Rule

*,

for

finding the fums of the powers of

the roots of an equation

L
x
x*
.
. .

ET

the

equation

be

ex
n ~~ l

Ax

Ix*

4 Kx ~Lx + M
1

~ -f ^
K

d x &c. ==
2

"a-Xx ^X
or,
.

o,

- Cx n i

7 __ i

known that A a + b + + d + &c. = ab + ac + ad + be + bd + cd + &c. C = #&: + *W + bed + &c. D = + &c.

It
is

tfjfttf

the parts or #n#.s of the coefficients A, B, C,

D, &c. being of
that
is,

i 3

2,

3,

4, C5V.

dimexfions

containing as

many

roots ov factors as

there are terms of. the equation preceding them,

reflectively.

* 9ee
15

Arith.

Univtrf.

pag.

157.

And

Chap.

II,

173

Of

this

P*r/a

CASE


Chap. 12.

ALGEBRA.
CASE
I.
;z,

287

Let r be an index equal to

fubftituting fucceflively ^, b y

or greater than

#, then, multiplying the equation

c,

by x r ~ *, and d, &c. for x,

you obtain
<r

. .

^f*r

...
.

- +Rr-*
I

Z./r-"* 1
l

+ Ma
r

Cars r ~*
r

.7_
j

- + h' Ab
r
r

Bb

-*Cb -i.
Mb'-*
2

cr

yfrby

Lb -+* +
1

...
&c.

+ 5t i> "+ +
r
1

CV
M<*-*

H.
this

Whence,
Theorem

tranfpofition
that, in

and addition,
cafe,

refults,

this

"

the

fum
r*

of the powers of the roots, of the exponent

is

equal to the

fum of

their

powers of the ex-

ponent r
plied

multiplied by

A> minus the fum

of their powers of the exponent r

multi-

by B,

It

the

fum of

thofe of the exponent

fo

3 multiplied

by C, and

on."

remains to find the fums of the powers of

the roots,

when

the exponents are

lefs

than n

the exponent of the equation.

CASE
If r
is lefs

II.

than

n,

and

be the coefficient,
;

in the equation,

of the dimsnftons r

that

is,

if

be taken fo that the number of terms precedit

ing

in the

equation be equal to

r, or the

number

288

^Treatise
its

&f

Part it

bcr of faffors in

parts abcdefgh, abcdefgi,

&c*

equal to r, then the

in the following

Theorem may be

exprcfTed

manner.
h r -r c r br

ar
r
1
I

+d +
r

Sec.

+b ~ = ^ + ^~ >X +i + &c. J
J
r
1

~ A ^~

z
2

+ b ~i
r

i' &c.j
2 i

>x J5+r-3
4- ^~~ 3

Jx'C-'-

r
r

&c.

H.

The
flrated
^

cafe
for,

when

.
.
.

is

eafily

demon-

dividing the equation by x 9

we

have

# x

Ax
n

+ Bx

-i

LA

= c.
o*

"Whence
j/t-j

Aa* Ab ~
n

+ Ba*-i ....- L ]
4-

M=
-7-

^"~3

....

L+ =
&c.

o*

&c.
and (becaufe

=
1

we

fliall

have a*-

+ T% 1
an

^
fi

+ T T "^ + J*-* + c"- + fife

i~

"J
3

"*~

-f

#""""+">

&c.J

+&c

J
L.

...... + 1 x

When

Chap. 12,

When
rived
z

ALGEBRA. demonftration r = #
2, the
%

289
is

de-

= A iB
By
x*
.
.

from hence, that a

+b

-\-

-f-

+ &c*

(fag. 142. J as follows.

Ax*- + Bx*-* Cx*-? ~p X#* * + M

1

32. transform the given equation, viz*

.

:-ljP

the equation

the roots a,

(3,

^,

&c. of which new equato the recipro-

tion fhall be rcfpectively equal

cals

abed
,

-,

-7,

&c.

of the roots of the

original

equation.

Divide
in
c,

now

the original equation

by *% and

the quotient fubftitute for x the roots a, b y

d9 &c.

08*

Aas
la 4-

fuccefnvely, fo fhall

4- Ba ~*
n

Ca s
n

you have

h ~r

^-s z

AC*~3 + Be** Cc ~$ ... 1 0~

L M t v -/,+---- + _
\

yfz-8-3 ~|- 2fc"-4

...

>

o>

J
and
will

&c.
n

Add

all

thefe equations
its

together,

for

fubftitute

value

r,

and

it

be
a*

'290

^Treatise

of

Pmlh

&C.

&C.J

+ &C.

~ b Ll+nK-xL+Mx &c.\
c

But by the principle adduced from pag* 142

*s

(3*

L 2K + ?* + &c. = -^ - -^
7,

wherefore,
it

by multiplication and low that

tranfpofition,

will fol-

2K-xL + Mx\%P
Which
ceding, there remains

f=o. [+&C.J

equation being fubtra&ed from the pre-

X~ b
c

>Xl + rxK=o.

Which was

to be proved.

But

to

mew

it

univ erf ally %

:

we may

ufe the'

following

Lemma

si

That

Chap. 12.
cc

ALGEBRA.
if

291
dimen-

That

A
G G

is

the coefficient of one

fion, or

the coefficient of the fecond term, in

an equation,
efficient

any other
it
;

coefficient,

H the coif like wife

next after

the difference of the di-

menfions of

and

A
in

being r

A X G'
root, as
will

reprefent the

the product
2
<3
,

AX G
or b z
,

A xG AG rH."
is

all thofe terms of which the fquare of any or c\ &c. is found ; then

fum of

This

a particular cafe of Prop. VI. concern-

ing the impoffible roots in Phil. Tranf.

N 408

which, by continuing the Table of Equations

in pag, 140,

and obferving how the coefficients

are formed,

may be

thus demonftrated.

Let
as

the coefficient of a term of the equation,

Sec.)

and,

+ bedt + bedf be multiplied by A(=za + b + + d + &c.) in the product A x D fetting afide the
(== abed

+ abce + abef,
9

&c.

all

terms,

A x D'

in

which a\ b\ c\ &c. are

terms of
the

found,

any one of the remaining terms will

arife as often as there are factors in the

the

following

coefficient
:

E.

Thus
it is

term

abed: will arife five times

becaufe

of any one of the

*z,

five roots (or

made up terms of A)

b9

e,

d,

e,

multiplied into the other four that

make

a term of

the like

is

true of every

other term, as abedf^ bedef, &c. each of which

will arife five times in the product

A X D.
abedf

And

the

fum of making up

thefe

terms abode
coefficient

+
it

&c.

the

follows that

Ax

292

A
the

Treatise
or

of

PartlL.

AxDA'xH=: 5 E,
And
flectively.

A'xD'= AD 5 K
two

fame holds of any

G, By whofe dimenfions are r

coefficients

and r

re-

To

apply

this to the prefent

purpofe,

it is-

to

be obferved

that,

in

each of the coefficients

lafi

By C,

Dr

&c
all

except the

M,

which

is

the product of

may
is

diftinguifh

bers, in

the roots a^ k, e, d^ &c. we two feveral portions or memone of which any particular root, as #,
the whole remaining porcoefficient,

contained,, but in

tion

of

the
is

fame

that
if,

particular
for

root (a)

wholly abfent.

Now
a,
is

bre-

vity's fake,
efficient

we denote

that portion of any co-

wherein any root, as

contained,

by annexing the fymboL of the

the fign

faid root

with

in an uncus to the fymbol, as

G, of

that coefficient (thus

C)

and

if

we denote

the remaining portion of the fame coefficient,,

from which the fame root a k by annexing the fymbel of the

fign

totally abfent

faid root

with the

in

an uncus to the fymbol

of the fame

coefficient (thus

')

it

will appear that (if

be any coefficient and

the following co-

efficient)

G=G VI + G&c.

ana r

= aG

K
,

&c.
Divide*

Chap. 12.
Divide

ALGEBRA,
the equation propofed

293
by x n ~ r 9

now
I

and

it

will

become
1

~ x r Jx r
.
.

- Cx'S + Ex * Cx K & + Gx H + a* ,
c4

1
1

= q

in

which fubftituting

&c.

fucccefiively

for #,

we obtain
r
1

ar

^ - + Bar"* Ca' Bar-* Lars Adr-*

r

* AV-* + Bb
.
.

-*

CI//.x

-j- vji/

"
r
1

z.3

.__
"j

Ac -

Bc r -* z
C

CV3
7
-

~r J

But, by the notation here ufcd, and explained

as above,

Ga = aG {+a +aG

-h=
1

~a - aG - a ~rt- a
(

~
_

7 {T

+-*.-

Whence

294

-#

Treatise
Whence

of

Part II,

~
Gc>

~P
4~

~r

Hi ~
&C.
c

C*
c

"-r

= cG ^
K

And

the

fum of
&c.

thefe

aG

{+a

^ +b

cG^ +

(by this notation)

/xG' =

(by the lemma)

xG rH.
Compare
this Iaft conclufion

with that which

followed from dividing the propofed equation

by x n r, and fubftituting for x c, &c. and you will have

the roots a> b^

+a +
which was
&c.

to be demonftrated.

From

Chap.

12.
thefe

ALGEBRA.
two Theorems
follows.
Sir Ifaac
iiluftrate the
:

295
Newton's

From

Rule manifeftly
But,, to

reafoning here ufed by

to take

fome examples fuppofe r == 3, then we are C for H, becaufe three terms only pre-

cede

in the equation x n

Ax""

-{-

Bx n~~ z

Cx 3

~f-

&V. == o

and we are to prove that

&c.

That

this

may

appear, obferve

that

a-\-b-\~c-\-d\&c.

h-\-c-^r d-\-tc.

*x * + '+<*+&?'. ^X^-f--|-^+&V.
2
/

i x7^+7 + ^. ^. = (becaufe^ 5'=

X ab-\-ac+ad+&c.
-\-bxab-\-bc-\-bd-\-Zc.

<z

+ cxac-\-bc-\-dc + &c. + dx ad+ bd+cd+ &c. + 6?r.) = a' + P+c' + lF + &c. x A AE (by the Lemma) = + ** + * + + &.
1
tf

rf

x^ 4B+3C
In like manner,
*'
5

a*+

b+-{-

c++d

+ * + + i* + &fr- X a + b + c + d+ &c. a *jr b*Jr + T + &c. X fl^+tff+^+^+^+fi + &V. + a -xbc + bd+cd-{- &c. + x^-J-^+rJ
f
3

+ &c. =

+ fcfh

296
c
z

Treatise

of
1

Part

II.

+ && + X ab-\-ad+bd+ lie. + d xab+ac+bc + &c. + &c. = a* + i> + r + fr + &c. x A <? 4. + ^~X~7 4- x ^ 4- AC ;=
2

fcfr.

End

&f the

Second Part,

A TREA-

TREATISE
O F

ALGEBRA.
PART
Of
III.

the Application cf Algebra

and

Geometry to each other.

CHAP.
Of
the Relation

I.

between the equations

of Curve Lines and the figure of thofe

Curves, in general.

i.

5Cfc3L^(N

the

two

firfb

parts

we

confidered

^
It

Algebra
metr y
'

as

independent of Geoits

)QRO(

anc^ demonrtrated

ope*

rations

from

its

own

principles.

remains that

we now

explain the ufe of Al5

gebra in the refolution of geometrical problems

or


298
or

^Treatise

of

Part III.
figures
;

reafoning about

geometrical

and

the ufe of geometrical lines and figures in the

refolution

of equations.

courfe of thefe fciences

tenfive

The mutual interhas produced many exthe

chief 'of

and

beautiful

Theories,

which we
equations.

fhall

endeavour to explain, beginning

lines

with the relation betwixt curve

and

their

2.

We

are

now
lines
\

to

confider quantities as

reprefented by

given
line.

line,

known quantity by a and an unknown by an undetermined

a
fufficient that
it

But

as

it is

be indetermined
to

on one fide, we may fuppofe one extremity be known.

P
1

A
i

F
i

P
3

B
1

AB, whofe extremities A and may reprefent a given B quantity while AP, whofe extremity P is undetermined, may reprefent an undetermined quantity. A lefTer undetermined quantity may be reprefented by AP, taking P nearer to A and, if you fuppofe P to move towards A, then will AP, fucceffively, reprefent ail quantities
Thus
the line
are both determined,
:

lefs

than the

firft

AP
it
9

and

after

has coin-

cided with A, if
tion to the place

proceed in the fame directhen will

hp

reprefent a nepofitive.

gative quantity, if

AP

was fuppofed

If

Chap.
If

i.

ALGEBRA.
reprefent x 9

299
then will
if

AP

Ap

reprefent
reprefent
(

x
(+

and

Ap

= AP,

and

for the

fame reafon,

AB
%
fent

a>)

then will

Ab (= AB)
if

re-

prefent
3.

a).

After the fame manner,

jy,

PM

repre-

and you take Pm, the continuation

of

PM
Pm

on the other
reprefent

will

fide,
:

equal to

PM,
,

then

for,

by fuppoling
decreafes
;

M to
when

move towards

P, the line

PM
it

M comes
4.

to P, then

PM vanifhes
m
9

and

after

M
ex-

has paiTed P, towards

becomes negative.

In Algebra, the root of an equation,

it
\

when

is

an impoffible quantity, has

it

its

preffion

but in Geometry,

has

none.

In

Algebra
there
is

you obtain a general

an expreffion, in
,

folution,

and

all cafes,

of the thing

required
preffion
rather,

only, within certain bounds, that exreprefents

an imaginary quantity,

or

"

is

the fymbol of an

operation which 9

in that cafe, cannot be performed j"

and ferves
only

300
only to
the
limits

A
mew

Treatise
it
is

of

Part III.

the genefis of the quantity, and

poiTible.

within which

In the geometrical refolution of a queftion, the thing required

cafes
is

exhibited

only in thofe

when
;

the queftion admits of a real fblu-

tion

and,

beyond thofe
in finding

limits,

no folution

appears.

So

the interfections of a
line, if

given

circle

and a ftraight

you determine

them by an equation, you will find two general expreffions for the diftances

of the points of

interfection

from the perpendicular drawn from But, geometrithe center on the given line.
cally, thofe

interferons will be exhibited only

when
center
circle.

the diftance of the ftraight line from the

is

lefs

than the

radius

of the

given

5.

" When

in

an equation there are two

undetermined quantities, x and y, then for each particular value of x, there may be as many
values ,of
tion."

as

it

has dimenfions in that equa-

Mr mI
i

'

Mi

A
111

So

Chap.
So

i.

ALGEBRA.
AP
a?,

301

that, if

(a part

of the indefinite line

AE)

reprefent

and the perpendiculars

PM
then

reprefent the

correfponding values of y,

there will be as
ties

many
y

points (M,) the extremi-

of thefe perpendiculars or ordinates, as there

in the equation.

are dimenfions of

And

the
,

values of
arifing

PM

will

be the roots of the equation

for

by fubftituting

its

particular value

AP
on

in any cafe.
it

From which
is

appears,

how, when an equati-

given,

the points

as

you may determine as many of you pleafe, and draw the line
all

that mall pafs through

is

thefe points

" which

called the locus of the equation."

6.

When

any equation involving two un(x and y)

is

known
equation

quantities

propofed, then

fubftituting for x any particular value

AP,

if

the

that arifes

has

all

its

roots pofitive,

the points
if

will

lie

on one

fide

of

AE
of

but

any of them are found negative, then thefe

are to be fet off

on the other

fide

AE

to-

wards m.
If,

for x,

which

is

fuppofed undetermined,

you you
locus

fubftitute a negative quantity, as

will find the points
is

Ap, then
:

M,

m, as before
all the
all

and the

not compleat
it

till

points

are taken in, that

may fhew

the values of

y correfponding

to all the pofTible values of x.

cafe,

If,

in

any

one of the values of y

vanilh,

302

A
If
it

Treatise
M.

of

PartUL

vanifh, then the point

coincides with P, and

the locus meets with

AE

in that point."

"
then

one of the values of y becomes infinite, (hews that the curve has an infinite arc
that
cafe,

and, in

the line

PM

becomes" an
it

efymptote to the curve, or touches

finite diftance,"
if
is

at

an in-

AP

is itfelf finite.

"

If,

when x
as

fuppofed infinitely great, a

value of y vanifh, then the curve approaches to

AE
"
fo

produced

an afymptote."

If any values of

y become

impojfible^

then

many points M vanim." 7. From what has been faid it appears, that when an equation is propofed involving two undetermined quantities (x andjyj " there may be
as

many interferons of
y

the curve that

line

is

the lo-

cus of the equation,

and of the

PM as there
;

are dimenfions of

in the equation

and

as

many
If

interfe&ions of the curve and the line

AE
the

as there are dimenfions

of x in the equation."
line

you draw any other

LM meeting

tul
fame curve
in

M,

and the

line

AE

in the

given

Chap.

i.

ALGEBRA,
ALM.
rife

303
== u 9 and

given angle

Suppofe

LM

AL = z
zy
fhall

-,

" then

the equation involving u

and

not

to

more dimenfions than y and

equation, or, than the

x had

in the propofed

fum of

their dimenfions in

any of
the

its

terms."

For, fince the angles

are given,
it

PLM, MPL, PML,

fines

follows

that,

of thefe
/,

angles being fuppofed to one another as

m, n 9

PM ML
:

(y

u)

and confequently
:

~ m

and that
,

PL

= m

and

ML n m\ fo that * = AP(=AL PL)=rz~- f?. m

PL
:

Subftitute, for
thefe values
(fince u

y and

#, in the

propofed equation

and z

and

it

is

obvious

and z are of one dimenfion only in the

that in the equation

values of
will

y and x) arife, z and u

fions

which will not have more dimenthan the higheft dimenfion of x and y in
dimenfions
are

the
their

propofed equation, or the higheft fum of

taken
together in the terms
:

where they
tc

both found

and confequently,

LM

curve

will

drawn any where in the plane of the not meet it in more points than
units
in the higheft

there are

dimenfion of x

or y 9 or in the higheft fum of their dimenfions, in the terms where both are found." Now
the dimenfion of the equation or curve being

denominated from the higheft dimenfion of x

ovy

in

it,

or from the

fum of

their

dimenfions

where

304
number of

Treatise
\

of

Part III.
that

where they are moft

with any fcraight

we conclude,

"

the

points in which the curve can meet

line, is

equal to the

number

that expreffes the dimenfion of the curve."

It

appears alfo from this article, how,

is

when

an equation of a curve
relation of the ordinate

given exprefiing the

PM

and

abfcifTe

AP,

you may transform

abfcifTe

it,

fo as to exprefs the rela-

tion between any other ordinate

ML
y
.

and the
its

AL, by
for

lubftituting

for

value

m
-

and

its

value z

Or, if you would have the abfcifle begin at e 9 fubftitutt any other point B, fuppoiing AB

=
+

for

x not z

nu
,

nu

but z

e.

8. Thofe curve lines that can be defcribed by the refolution of equations, the relation of

whofe ordinates

PM

and

abfcifies

AP

can be

exprefTed by an equation involving nothing but

determined quantities befides thefe ordinates and

abfciiTes,

are

called

tc

geometrical or

algehralc

curves."

They

are divided into

orders

according

to

the dimenfions of their

equations, or

number

of points in which they can

line.

interfect a ftraight

The
firfl

firelight

lines

themfelves

conftitute

the

order

of

lines

and when

the equation

exprefling the relation of x and

is

of one

dimenfion

Chap.

i.

A L G E BRA.

305
all

dimenfion only, the points

M muft be

found

in a ftraight line conflituting a given angle with

AE.
Suppofe, for example, that the equation given
is

ay

bx cd =
*

o,

and that the

locus is re-

quired.

Since * =5 y

it

follows, that,

a
if

APM be-

ing a right angle,

you draw

AN

making the

angle
as

NAP
\

fuch that

its

cofine

be to

its

fine

a to b

and drawing and equal

dinates

PM,

AD parallel to the orto through D you

,

draw
fame
of

DF
fide

parallel to

AN, DF

will be the locus

required.

Where you are to take AD on the of the line AE, with PN, if bx and
they have contrary figns.

cd have the fame fign, but on the contrary fide

AE
9.

if

Thofe curves whofe equations are of

conftitute
firft

two dimenfions
lines,

the fecond order of

and the

kind of curves.

Their

in-

terferons

jo6
two, by

A
7.

Treatise
a ftraight
line

of

PartllL

terfections with

can never exceed

The

curves whofe equations are of three dilines,

menfions form the third order of

cond kind of curves
:

or fe-

and

their

interferons with

ftraight line can never exceed Three.

And,
deter-

after the

fame manner* the curves

are

mined
finity.

that belong to the higher orders, to in-

Some
number

curves,

if they

were completely de-

fcribed, could cut a ftraight line in an infinite

of points

but thefe belong to none

-,

of the orders we have mentioned

geometrical
tion betwixt their

they are not

or algebraic curves, for the relaordinates and abfchTes canfinite

not be expreffed by a
only
ordinates

equation involving

and

abfciffes

with determined

quantities.

10.

As

<c

the roots of an equation

impoffible always in pairs, fo the

become interfections of
in

the curve and

its

ordinate

PM

muft vanifh

pairs," if any of

them

vanifh.

Let
touch

PM
it

cut the curve in the points

parallel

and
to

m, and by moving
in the point

to

itfelf

come

then the two points

of interferon, of contact N.
itfelf,

M
If

and m, go into one point

PM

ftill

move on

parallel to

the

points of interferon will,

beyond

Chap.

i.

ALGEBRA.
;

307

N, become imaginary

as

the two roots of

an equation
ginary.

firft

become equal and then ima-

ti.
ders,

The
all

curves of the 3d, 5th, 7th or-

and

whole dimenfions are odd numbers,

at
leaft,

mult have,
equations

two

infinite

arcs

fince

whofe dimenfions are

real

odd numbers

have always one

root at lepft\ and confe-

quently, for every value of x, the equation by

which y
real root:

is

determined mud,
fo that as

at leaft,

have one

x (or

AP) may
it

be increafed
that

in

infinitum

on both

fides,

follows
fides,

muft go off
limit.

in infinitum

on both

without

Whereas,

in the curves

whofe dimenfions are

of their equations
it

even numbers,

as

the roots

may become

all

impoflibie,

follows that the

circle

figure of the curve

may

be like a

or oval
that

308
that
is

A Treatise
limited
it

of

Part III.

within certain bounds, beyond

which

cannot extend.

12.

When
is

two

roots of the equation

c;

by
the

which j
ordinate

determined become equal, either

PM

touches

the curve,"
cafe,

two

point.,

of interferon, in that
of contact* or,

going into a point

"

the point

M
its

is

zpunffmn du-

plex in the curve -" two of

:

arcs interfering

or, " feme oval that belongs each other there to that kind of curve becoming infinitely little in M, it vanifhes into what is called a pun5itim

conjugatum."
If,

in the equation,

y be fuppofed
diflances

= o,
is

then
deter-

"

the roots of the equation b; which x

will

mined,

give

the

of the points

where the curve meets

if

AE

from A."

And, "

two of

thofe roots be found equal, then either

"

the curve touches the line

AE;"

or,

AE

paffes

through a pimclu:n duplex

in the curve."

When

is

fuppofed == o,

if

one of the values

that cafe, paries

of x vanifn, " the curve,

in

through A."
duplex'
9

If

two

vanifh, then either

," or,
kC

"

AE

touches the curve in

is

a punftum

" As

a punffium. duplex

is

determined from the

a

equality of

two

roots,

fo

is

punftum

triplex

de-

termined from the equality of three roots.

13

few examples will make thefe obfer-

vations very plain.

defcribe the line that

Suppofe
is

it

is

required to

the U'cus of this equation,

Chap.
tion,

ALGEBRA. y = ax + ab or y ax ab~o.
i.
z
z
9
9

309
Since
in,

y ==

i *Jax-\-rfb'
value,
it

and

firice

^ and are given

variable quantities, if

you afiume

AP

x) of a
-\-

known
and

will

be eafy to find

V ax

ab\
9

fetting off

PM on one fide equal to vax+ ab

PM,
pcfitive value of

and

Pm
and

on the other equal to

the points

will

belong to the locus required.

And
The

for every

AP

you

will

thus obtain a point of the locus on each fide.

greater

AP (= x)
the greater.

is

taken, the greater does

the *lax-{- ab become, and consequently

PM and

Pm become
If

AP

be fuppofed infinitely great,

PM

and

Pm
go

will alfo

become

infinitely great

and conand from

sequently the locus has two infinite arcs that

off to an infinite diftance
If

from

AE

AD.
fo that

you fuppofe x
and

to vanifn,

y ==

v ab
'

y does not

vanifh in that cafe but paffes

through
a

d>

taking

AD

and

Ad
b.

V ab

mean
If

proportional betwixt a and

you now fuppofe that the point P moves to the other fide of A, then you mud, in the equation, fuppofe x to become negative, and
'

= + slab ax\
9

fo that
is lefs

will
b.

have two values

as before, while x

than

But

if

AB

and you fuppofe the point P to come to B, then

ab

= ax PM and

and y

= + si ab ax o.
;

That

is,

Pm
AE.

vanifh
If

and the curve there meets

the line

you fuppofe P to move from

310

^Treatise
ab,

of

Part

III.
b,

beyond B, then x becomes greater than

lb that

and ax greater than

ab

ax being

negative,

slab

ax

becomes

imaginary, and
-

the two values of y

is,

become imaginary

that

beyond B there are no ordinates that meet

limited in B.
is

the curve, and confequently, on that fide, the

curve

is

All this agrees very well with what

known

by orher methods,
tion
is
z

that
is

the curve whofe equaa parabola

is

= ax + ab,

whofe vertex
to a.

B, axis

BE, and parameter equal

b _ x, and

For
will

fmce
to a

BPz=
-,

PM ==j^
BN
:

if

BF
the
is

be equal

then the rectangle

(=zab +_dx)
is
it

be equal to

PMg

(==

;)

which

known

property of the parabola.

-

And

obvious,

that the figure of the parabola

is

fuch as

we

have

Chap.

i.

ALGEBRA.
this Jccus to
it

311

have determined

be from the con-

fideration of its equation.

14.
that
is

Let

be required to defcribe the line

= be +
1

the locus of this equation, xy be A- bx bx, or jy == J *.a-\-c-\-x

+ ay + cy
can meet

Here,

it is

plain,

the ordinate

PM

the curve in one point only, there being but

one value of y correfponding to each value of x.

When

= o,

then y rr

-
If

fo that the

curve

does not pafs through A.

increafe, then

x be fuppofed to

will increafe,

but will never be-

come
a
-f- c

equal to b 9 fince v 1

r^ x
-\- c -{-

and
If x

+x

is

always greater than

c -\- x. c

be fuppofed

infinite,

then the terms a and

v,

va-

rum compared with

and confequently
it

y~b
it

b; from which
b,

appears, that taking

parallel to

AD
If x

and drawing

GD

AE,
at

will be an afymptote9
infinite diftance.

and touch the curve

negative,

an

be

now
the

fuppofed

and

AP
fhall

be taken on
y ' 3= b
fide,

other fide of
j

A, then

xa

-f- c

and

if

x be taken, on that

c,

then fhall

y= b

x
c9

=o
AB

fo that
c.

the curve muft pafs through B, if

If x be fuppofed greater than

then will

become

negative, and the ordinate will

become
negative

312
negative and

A
lie

Treatise
on the other

of
of

Part III.

fide

AE,

till

becomes equal to a

c9

and then y

=b

x
5

P
P B

or

infinite

fo that if
will be

AK

be taken =. a

r,

the

ordinate

KL

an afymptote to the curve.

If # be taken greater than a-\-

than

AK,
\

then both

x and
fince

c,

or

a-\- c

x become
)

AP greater

negative

and confequently y (= b
pofitive
5

X
c
is

becomes

and

always
greater

Chap.
greater

i.

ALGEBRA. than x a
c, it

313

follows that

always greater than b or

the reft of the curve

KG,
in

y will be and confequently

the angle

lies

And,

xa

as

x increafes, fince the ratio of x

c
it

FGH.
c

to

approaches

ftill

nearer to a ratio of

equality,

follows that

PM

approaches to an
its

equality with

PN, and
fide alfo.
is

the curve to

afymptote

GH

on that

This curve
fince b

the

common

hyperbola

for

y x a + c + #, by adding ab to both fides bx a + c + x ~yxa + c + x + ab\ and b y X a c + x = ab; that NMxGN = GCx BC, which the property of the comX
c-\~

-\-

is,

is

mon

hyperbola.
the

And
locus

it

is

eafy to fee

how

the

figure of

we have been

confidering

agrees with the figure of the hyperbola.

15.

Let

it

be required to defcribe the locus

cy*
z

of the equation
2 fince v
!

xjf x -f bx*. x and y = -f J X v

!

Where,
,

it

rol-

lows that both

PM
to

and

Fm
C

muft be taken equal, on

*
-.

fides,

Jv
^f
c%
if

But that when x

r,

is

taken equal to
dicular to

AB

and

BK

be perpen-

AB,

then

BK

muft be an afymptote
r,
.

to the curve.

If x be fuppofed greater than

then c greater than x 2 -J- bx* tive, the fraction will become negative, x c> and its fquare root impoiTible. So that no part
or

AP

AB,

x being nega-

of

314
of the
locus

Treatise
P

of

Part III.
If x be
fide

can be found beyond B.

fuppofed negative, or

taken on the other

*
,

of A, then

= + *V

*
,

the fign of x l
fign of bx
z
,

and # being changed, but not the

became the fquare of

while x
real
is

a negative
its

is

the fame as
is

the fquare of a pofitive, but

lefs
;

cube

negative

than

the values of

will

be

and equal

but

if x ==

then the values

of y vanifh, becaufe,

in that cafe,

confequently, if
will pafs

be taken == , the curve through D, and there touch the or-

AD

dinate.

If x be taken greater than b> then

will

+ bx

c+x

become
is

imaginary^ fo that no part of the

curve

found beyond D.

If you fuppofe

o,

then will x*

+ bx = o
z

be

Chap.

ALGEBRA. be an equation whofe

i.

315
h 9 o, o,
-

roots are

from

which

it

appears that the curve paries twice

in

through the point A, and has,

duplex..

A, apunclum

This

locus

is

a line of the third order,

it

BK is

its

afymptote, and

has a nodus betwixt

and D.
If

you fuppofe b
z

fo that cy

to vanifh in the equation,

xy

=1

x\ then

will

and

coin-

cide,

and the nodus vanifh, and the curve will

in the point

have

A a cufpis,

the

two

arcs

AM
and

3 16

A
Km
this
is

Treatise

of

Partlll.

and

touching one another in that point.

the fame curve which

And
cients

by the anline

was

called zhc cijfoid of Diodes, the

AB
and
nate

being the diameter of the generating circle,

BK

the afymptote.
if

For,

BR

be equal to

AP, and

the ordi-

RN

be raifed meeting the

it

circle in

N, and

AN
in

be drawn,

will cut the perpendicular

PM
be a
:

a point of the ciflbid.

:

So that
:

if
:

M
z

point in the cnToid,

AP PM
:

AR RN

VAR VBR
:

VBP

v/AP, and confequently,

that
i$,

BP X PM# = AP cub.
which
If,
is

c-x xy = *

3
:

the equation the locus of which was re-

quired.
inftead of fuppofing b pofitive, or equal

to nothing,

we now
cy
z

equation will be
will

# bx
xy
z
3

fuppofe

it

negative,
z
,

the

the curve

pafs

through D,

as

before,

and taking
be
its

AB

c BK will
:

a-

fymptote

it

will

have a

punffum conjugatum in A,
becaufe

when y

vanifhes,

two values

of x vanifh,
e-

and the third becomes

qual

to b or

AD.
befides

The
this

whole curve,
point

A, lies between DQ_ and BK. Thefe are demonflrated after the fame manner as in the
firft cafe.

16.

Chap.

i.

ALGEBRA.
is

317
y

16. If an equation

propofed, as
is

+ bx
two

n ~~ J

cx

z &c.
^

= ax

and

an even

num-

ber, then will the

locus

of the equation have

fide

infinite arcs lying

on the fame

of

AE.

For, if x become
negative, x n will

whether pofitive or be pofitive, and ax n have the

infinite,

fame fign
infinitely

in either cafe-,

and

as

ax n becomes

greater than the other terms #*

it

cx n

z
,

&c.

follows that the infinite values of

cafes
;

will

have the fame fign in thefe

and

confequentlyy the two infinite arcs of the curve

will lie

on the fame
if

fide

of AE.

But

n be an odd number, then

when x
is

is

negative, x n will be negative, and ax n will have

the contrary fign to
tive
cafe,
-,

what

it

has

when x

pofi-

and therefore the two

will lie

infinite arcs, in this

on

different fides of

AE, and

tend

towards parts directly oppofite.

Thus

the

locus of the equation ay

=x

is

the parabola.

is

the vertex,

AE

is

the tan-

E
gent at the vertex;
lie

and the two

fide

infinite

arcs

manifeftly

on the fame

of

AE.
But

3 i8

A
locus

Treatise
is

of

Part III
#',

But the

of the equation a*y ==

where
its

the index of x

an odd number, has

two

m/k-

arcs

on

different fides
as.

of

AE,

tending towards

oppofite parts,
is

AMK,

and Amk. and

This curve
is

called the cubical parabola,

a line of

the third order.

The

locus

of the equation

ay
y

x*
;

figure like the

common

parabola
is

is of a and " all

thofe loci, in whole equations

of one di:

mension, x of an even number of dimenfions

But thole

loci

are like the cubical parabola, in

whofe equations y is of one dimenfiqn only, and x of an odd number of dimenfions." And this

Rule

is

even true of the locus of the equation

= x,

which

is
;

a ftraight line cutting

AE

in as

an angle of 45
pofite parts,

which manifeftly goes off

the cubical parabola does to infinity, towards op-

and on

different fides of

AE.

Chap.

i.

ALGEBRA.
n -=.
%

17. If the locus of the equation yx

is

319 an * 1

required

If n
tive,

is

an odd number, then when x

j

is

pofi-

y ==

but when #

is

negative, then

+
fo that this

curve rnufl

all lie

in
(as

the vertically oppqfite angles

the

KAE,

F*AE,

common

hyperbola

:)

FK, E?, being afymp-

totes.

But

if

is

an even number, then y

is

always

pofitive,

whether x be pofitive or negative, bein

this

caufe x n ,

cafe, is

always pofitive

and

there-

32o

^Treatise
muft
all
lie

of

Part III.

therefore the curve

in the

two ad-

jacent angles

KAE

and KA^, and have

AK

and

AE

for

its

two afymptotes.
7-

3.

Let the equation given be a

b\

~x y
t

z
\

fo

that^ 3=

+ */a
If x

~x X

= o,
the
is

then

becomes
therefore

infinite,

and

ordi-

M
M
P

nate at
ptote

an afymcurve.

to

the

IfAB

= J,

and

P be

taken betwixt

A and
lie

B^iB 771 m

B, then lhall

PM and
AP.
If

Pm
on
#

be equal, and

different fides of

the abfciffe

= ,

then the two

values

of y vanifh,
becaufe

Chap.

i.

ALGEBRA.
b = o;
AP
be taken greater than

321

becaufe x
pafTes plex.
fhali

and consequently, the curve

through B, and has there a puntlum duIf

AB,

then

there be two values of y,

figns, that

as before,

having

contrary

value which was pofitive

negative, and the nepofitive.

before being

now become
being

gative

value

become

AD
And
az

be taken

a
is

and

P comes

to

But if D, then the

two values of y

vanifh,

becaufe sla z
greater

x =
z

.0.

x
If

if
z

AP

taken

than

AD,

then

impqffible

becomes negative, and the value of y and therefore, the curve does not go :

beyond D.
x

now be fuppofed
z

negative,

we mail

find

y + sic? x
thefe

+ x -r x.

If x vanifh, both

values of

quently,

y become infinite, and confethe curve has two infinite arcs, on

of the cfymptote

each
it is

fide

AK.

If x increafe,

plain

diminiihes, and if x becomes

=a

y vaniiiies, and confequently the curve parTes be taken AD, on the opthrough E, if

AE

pofite fide.

If x be fuppofed greater than a, then

impcjjible
,

y becomes

and no part of rhe curve

can be found beyond E.

choid of the ancients.
If

This curve

is

the con-

a =-

b, it will

have a

cufpis in

B, the nodus
if

betwixt B and

vanifhing.

And

is

lefs

than

by

the point

will

become a punflum

con-

jugatum.

From

322

A
des

Treatise

of

Partlll

From what
Royak
infinite

has been faid an error

may be

corrected of an Author. in the

Sciences,

who

gives

Memoir es de FAcad. this curve no

arcs,

but only a double nodus. of the fame kind

Treatife,

Some
be cor-

other
rected
faid.

errors
in

may

that

from what we have

19,

If the propofed equation can be refolv-

ed into

two equations of lower dimenfions,

without affecting either y or x with any radical fign, then the locus mall confift of the two loci

of thofe

inferior equations.

the equation

Thus

ixy

by

xz

bx == o

the locus of
is

found to be two ftraight

lines

cutting the ab-

fcifTe

AE
is

in angles of

45%

in the points
b>

A
o,

and
and

B, whofe diftance
tion

AB =

becaufe that equa-

x+
be

refolved into thefe

b

two y

x=

= o.
three

After the fame manner, fome cubic equations

can be refolved into

fimple

equations,
lines
;

and then the

locus

is

three ftraight

or

may

refolved

into

a quadratic and fimple

equation %

Chap.

i.

ALGEBRA.
is

323
and

equation, and then the locus

a ftraight
line.

a conic feclion

In general, " the curves of the fuperior or*

ders include ders
;

all

the curves of the inferior oris

and whatever
is

demonftrated generally
true of the
inferior

of any one order,

orders."

alfo

80, for example, any general proper-

ty of the conic feclions holds true of

lines as well as of a conic fedlion.

two

ftraight

Particularly

that
lels

"

the rectangles of the fegments of paralwill

bounded by them,
of the
lines

be always to one
general proper-

another in a given ratio."

ties

The

of the third order are true of

three ftraight lines, or of any one ftraight line

and a conic
perties

fection.

And,

as

the general prolines

of the higher orders of

defcend
there
is

alfo to thofe

of the inferior orders,

fo

fcarce

any property of the

of which

inferior orders,

but

has an analogy to fome property of the higher

orders
;

it

is

but a particular cafe or

inftance.

And

hence, the properties of the into the difcovery of thofe of

ferior orders

lead

the fuperior orders *.

20.

We

have fhewed how to judge of the

figure of a locus

from the confideration of

a locus
is

its

equation.

And when

to be defcribed

exactly, for every value of x

you muft, by the

Rules

refolution of equations* according to the

See the

Appendix.

in

324
in

^Treatise
II.

of

PartHL
y9

Part

find the correfponding values of

and determine from thefe values the points of

the
locus.

But there

are

geometrical

conftrutlions

by

which the roots of equations can be determined more commodioufly for this purpofe. And, as by
thefe conftru&ions

we

defcribe the

loci

of the

equations, fo reciprocally

when

loci

are defcrib-

ed, they are ufeful in determining the roots of

equations

both which

(hall

be explained in the
mail give an ac-

following Chapter.

Then we
loci

count of the moil general and fimple methods

of defcribing thefe
tion

of angles
;

and

lines,

by the mechanical mowhofe interferons

trace the curve

or of conftructing them by

finding
points.

geometrically

any

number

of

their

CHAP.

i.

ALGEBRA,

325

CHAP.
Of

II.

the Gonftrudion of Quadratic

Equations ; and of the Properties

of the Lines of the fecond order,

21.

rjr\HE
JL
having
this

genera] equation expreffing the

nature of the lines of the fecond

all
its

Order,
will

terms

and

coefficients*

be of

form;

+ +
d,

axy
by

+ * + dx + i
c

Where
tities

h
given, as
it

a, b,

e,

reprefent any given quan-

with their proper figns prefixed to them.

is

If a quadratic equation
-f-

q = o,

and, by comparing

py with the preced-

ing, if

you take the quantities

-\-

a, b, r, d, e y
x

and
9

x fuch that ax

b=zp, and
it

cx

-f dx

+e=q
;

then will the values of y in the

equal to the values of
if the locus

firft

equation be

in the

fecond

and
firft

be defcribed belonging to the

equation, the

two values of the ordinate when

ax

+ b z=.p
And
as

and cx x

two roots of the equation y x

+ dx + e == q will be the + py + q = o.
t

four of the given quantities a> b 9

c^

d$ e,

may

be

taken at pleafure, and the

fifth*

with

326

A
be
ftill

Treatise
/>,

of

Part

III.

wich the abfchTe r, determined, fo that ax

may

equal to

and cx z

+b + dx + e =2 q
;

hence there are innumerable ways of conftructing the fame equation.

But thofe
of

loci

are

to

be preferred which are defcribed moft eafily;

and therefore, the

is

circle,

all

conic feclions,

to be preferred for the refolution of quadratic

equations.

2i. Let

AB

be perpendicular to

AE, and

upon

AB defcribe the femicircle BMMA. If AP be fuppoled equal to x, AB s= #, and PM =zy, then making MR, MR, perpendiculars to the
diameter

AB,

fince

AR=y, RB ay, RM
ay

AR x RB = RMj,
z=z

and

x,

it

follows that

a yx-y =

x*,

andjy a
if

an

o.
z

equation y q be propofed to be refolved,

its

py + = o,

And,

roots will be the

ordinate to the circle,

PM

and

PM,

to

its

tangent

AE,

ifar=zp,

zndx z =:q:

becaufe then the equation

z of the circle y z ay -f x o, will be changed into the propofed equa~y z

tion

py +

= o.
this conftruclion for find-

We

have therefore

ing the roots of the quadratic equation y z defcribe a q =r o ; take AB /, and on

py

AB

femi-

Chap.

2.

ALGEBRA.
;

327
AB,

fcmicircle

then
take

raife

AE

perpendicular to
that
is,

and on

it

AP = y^,
1

mean pro-

portional between

draw

PM

parallel

and q (by 13 El 6.) then to AB, meeting the femilines

circle in

M, M,

and the

PM,

PM
"

(hall

be

the roots of the propofed equation.

It

appears
,

from

the

conftruclion

that

if

or

s/q

= ipy
in

then

AP = -JAB,
N,

and the
the two

ordinate
roots

PN

touches the curve in

that cafe,

PM, PM,

becoming equal

to one another and to

If

PN.
is,

AP
*Jq

be taken greater than -JAB, that

is

when
ip
as
1
,

greater than

the ordinates
roots of the

do

ip 9 or q greater than not meet the circle, and

:

the

equation become imaginary

in

we demonftrated,
II,

another

manner,

in

Part

23. The

roots of the fame equation

may

be otherwife thus determined.

Take
to

AB

y/q,

and

raife

BD

perpendicular

AB
-ip,

from

as a center with a radius equal

to

defcribe a circle meeting

z

the two roots of the equation y fhall be CB, and CB.

AC +

AC
ip

py + =
"

BD

in

C, then
q

o,

For

thefe

roots

are

^ip

q<>

and

ip^ip'-q,

and

At=ip, C6=*/AC -CB a

thefe roots are

\Z^p*
AC +
CB.

j,

and confequently

The

328

Treatise
%

of

Part III.

The

roots of the equation

y +py -f q

= o are

AC + CB;
manner.
24. q =z

as

is

demonftrated in the fame

The

roots of the equation

py
;

5-

are determined

by

this construction.

Take
the

AB

= *p,

BC Q ^,

draw

AC

and

two

roots fhall be

AB + AC.

If the fe-

lt/ 7

AB + AC.

cond term

is

pofitive, then

the roots mail be

And

Chap.

2.
all

ALGEBRA.
four forms,

329

And
to thefe

quadratic equations being reducible

f py + 4 = * y*+py q = o,
y'py z y +py
it

q^
be

+$=
articles.

o>
all

follows, that they

may

eonftru&ed by

this

and the

lafl

two

25. By
be defcribed

thefe geometrical confl ructions, the

locus of any equation of two dimenfions

-,

may

fince,

by

their

means, the values

of y that correfpond to any given value of x may be determined. But if we demonftrate

that thefe
loci

are

always

conic feclions^

then

they

may more

eafily

be defcribed by the me-

thods that are already

thefe curves.

known
we

for

defcribing

In order to prove this,

(hall

enquire what

equations belong to the different conic fe"ions%

and,

as

it

will

appear that there

is

no equation
are loci

of two dimenfions but mnft belong to one or

other of them,
it

will follow that they

of

all

equations of two dimenfions.

26.
line

Let

CML

be a parabola
;

AE
be

any
re-

drawn

in the

fame plane

and

let it

quired to find the equation expreiTing the relation

betwixt the ordinate

PM

forming any
abfciiTe

given

angle with

AE,

and the

AP

begin-

330
beginning at

Treatise

of

Part HI,

A any

given point in the line

AE,

Let

CF

be the diameter of the parabola

parallel

whofe ordinates are

to

PM.
in

Draw
;

AH
AD
lines

parallel to

CF

meeting

PM

and

parallel to

PM meeting CF in D.
will

Becaufe

the angles

HAE, APN, ANP,

:

are given, the

AP, PN, AN,

be in a given ratio to
,

each other
c
\

fuppofe them to be always as a,

let

AD == d, DC=ru;
b,

and feeing
likewife

AP (= x)

;PN.::
a
:

PN:=-#;
a
-#.

AP AN::
1

r,

or

AN =

r-

NG = Jy Z# d. a

But

CG

s=

PN DG
PC

Chap.

ALGEBRA. DC = AN DC = x
2.
e.

331

If

now

the pa-

rameter of the diameter

CF

be called p, then,

from the nature of the parabola,

and confequently, p

p x CG GMq x-

X x ey

from which

this

equation follows,

Whence,

if

any equation
e,

is

propofed, and fuch

values of a, h* c9 d,

can be affumed as to then the

make
locus

that equation

and

this coincide,

of that equation will be a parabola. of which

The

conftruction
this article.

may

be deduced from

27. In this general equation for the parabola, the coefficient of x

z

is

the fquare of half

the coefficient of xy,

tion
is

and,
this

" when any equaproperty, the locus

propofed that has

a

of

it is

parabola"

For, whatever coefficients

affect the three

lad terms, they

may

be

made

to

agree with the coefficients of the

the general equation,

laft

terms of

by afTuming proper values

of p,
It

c9

and

e.

appears

alfo,

that

"

if

the locus be a pa-

rabola,

x%

and the term xy be wanting, the term muft alfo be wanting." And, " if any
that

equation of two ditnenfions be propofed

wants

332

Treatise
xy and

of

PartllL
it

wants both the terms,

#%

may be
ellipfe is

always accommodated to a parabola.

28. The
any diameter,
26.

general equation for the

deduced from the property of the ordinates of

in

the fame manner-,

figure being

the conas in

duction of the
Only,

the

fame

in place

of the parabola,

Let

KML

be an

ellipfe

whofe diameter

is

KL,

having

its

ordinates parallel to

PM,

and

K
let

"T

P
ellipfe.

E
Suppofe

be the center of the

CL =

and the parameter of that diameter

z=p, then
in 26,

GM? CLj CGg

:

it.

But, as

GM =? y
y
:

x
x

d
\

and

CG=x ey
/

therefore,

d\

x
p

=#*

+
*

-x
a

whence

this

equation
,
;

7hdi
,

+
x
"*"

d1
2 /
2/ J

fee

= 0=
And


Chap.
2,

ALGEBRA.
is

333
by
af-

And
made

if

any equation

propofed that can be

to agree with this general equation,

#, b,
c,

fuming proper values of

d,

p and

e;

then the locus of that equation will be an

ellipfe.

29.

"In

the general equation for the

z
:

el-

z lipfe > the terms x and

the coefficient

y have the fame fign and z of x is always greater than the

coefficient

fquare of half the

of xy 9

becaufe

+ ^
a
z

is

2ta*

greater than . & az

And

although the &

term xy be wanting, yet the term x z

main,
its

mud
being

re-

coefficient,

in

that cafe,

On

which muft be always

the other hand,
if

real

and
is

pofitive.

an equation
z

propofed in

which the
wants
its

coefficient of x

exceeds the feu are of

or,

half the coefficient of xy

xy>

an equation that
of the

but has x

and

y\

fame

fign,

locus

muft be an

ellipfe"

30.

In the hyperbola, as

GMj

CGq

Chq

it

when

t is

a firft diameter, the

equation that

arifes will differ

from the equation

of the values
will

of the
of

ellipfe

only in the figns

CG^

and CL^, and confequently

have

this
z

form,
-

xy

-x z Pc
2tq
%

idy ^
*
1
l

x
P ce
at

+d
~
,

P*

O.

2 pe z

~~2~tl
If

334

-^

Treatise

of
then

Part III.

If / be a fee and diameter,

negative.

will

be

In this equation,
ficient

it is
z

manifeft that the coeflefs

;

of the term x

is

than the fquare of

and, that
.

half the coefficient of xy

when

the

term xy
gative.

is

wanting, the term x z muft be nereciprocally,

And,

"

if

an equation
is iefs
;

is

propofed where the coefficient of x z

the

than
or

fquare
is

of half the coefficient of xy

where xy

trary figns,

wanting and y z and x z have conthe locus of that equation muft be

an hyperbola ."

3^

Chap.
31.

2.

ALGEBRA.
PM
the general

335
its

The

equation of the hyperbola when

ordinates

are parallel to an afymptote does

not

come under
Let

equation of the

laft article.

CF

and
let

CL

be the afymptotes
to

of the hyperbola, and

PM be parallel

CL.

P
Then

CG X GM
e,

will

be equal to a given rectan-

gle (which fuppofe ga).

Then,

CG = DG
and con-

DC = x a
fequently

GM = y 'a x d\
c

X dx X
b

=gxa. whence
:

this equation,
b

xy

xx

eb
c

d
Where
found with xy
;

V
is

-*

will
if

ea

eda
"J

ia
c

c
[

"

0.

only one of the terms jy\

x\

can be

and where xy

be found with-

out either of thefe terms,

cide, that
is,

AE

and

AH

coin-

if

AE

parallel to the

afymptote
It

DF.

336
It appears
is

A
;

Treatise
this,

of

Part lit
an equation
z

from

that

"

if

propofed that either has xy the only term of

or,

two dimenfions

has xy and either x z or

befides, but not

both of them, the locus of the

equation mall be an hyperbola, one of whofe

afymptotes mall be parallel to
as
it
is
z

or x according
in the

or x

that

is

wanting

equa-

tion." 32. From all theie compared together, it follows, that " the locus of any equation of two*

dimenfwns

is

a conic fection."

For

if the

term xy

is

wanting

in the equation^
it,

z and but one of the terms y , x z is found in the locus fhall be a parabola ; by 27.

wanting, and x z 9 y z , have the fame But^ fign, then the locus is an ellipfe. 29.
If xy
is

when they have

3If xy
is

different figns,

it is

an hyperbola*
z

found

in

the equation, and #*,

are both wanting, or either of them,- the locus is

an hyperbola.
If

31. z both x and y are found

z

in

it,

having con-

trary
If y

figns, the locus

z

is

frill

an hyperbola.
figns, then, accord-

and x z have the fame

z

ing as the coefficient of x

lefs

is

greater, equals

or'

than the fquare of half the coefficient of xy*

cllipfe,

the locus mail be an

perbola,

parabola,

or hy-

27, 29, go.

In any cafe therefore the locus of the equation

is

fome

conic feclion*

3$-

Chap.

2.

ALGEBRA.
may
all

337

33. Thefe
directly

be demonftrated more of the fecond order in

II.

from the consideration of the general

equation of the lines

21.

For

it is

obvious that, by 25. Part

a W
the fecond term of that general equation

may be
,

exterminated by afluming z
it

= y + ax

and

will

be transformed into

"which,

by tranfpofing the

laft

term,

is

22

=
4

+
*
-\

Let

338
Let
if

A
MK
D
and

Treatise

of

PartllL
:

be the locus of the equation

and and
to

AH be drawn fo that HE
AD,
parallel to

be to

AE

as

a to

unit)

PM,

be

= ^b,

through

the line

DF

be drawn

parallel

AH, meeting PM in G, then fhall GM (= PM + PN + NG -y + iax + ib) =2 z. And if

AH =/,

then

Suppofe

DG = AN -fx. DG = #, and x =
and
the the relation of

Inftead.oftf
that
refults

fiMitute y,
will
this

equation

exprefs

GM
u
,

and

DG,

of

form,

z-

~ X *+ ^F-X
a
7-

Lc

al

2d

+ ib

,, z

e=:o.
ellipjis,

Which

will be

an hyperbola, -parabola, or

according as the term or negative.

equal to, or
cient of xy
,

-^-

is

pofitive, nothings

That
lefs

is,

according as
c.

is

greater,
coeffi-

than

But a was the

it

from which
ellipfe,

appears, that

" the

locus

is

an

parabola, or hyperbola, acis

cording as the coefficient of x 1

to, or lefs than

greater, equal

the fquare

of half the coeffi-

cient of xy."
It appears alfo,

that

"

if the

term xy be wantellipfe,

ing, or a

= o,

then the locus will be an

parabola, or hyperbola, according as the term ex*

is

pofitive,

nothing, or negative"

Hence

Chap.

2.

ALGEBRA.
z

339

-u
fide of

Hence likewife, if the term x be wanting, and the term xy not wanting, then the term
1
-

being pofitive (becaufe

- is

always

pofitive,

whatever a or

be)

<c

the locus muft be

an hyperbola."
Note, That part of the figure, on the other

AE, which

anfwers to the cafe

is marked with fmall letters, when the coefficient of y 9 in

the general equation, viz. ax

+b
be

is

negative.

$ 34.

The

lines

of the fecond order have fome

general properties which

may

demon ftrated

from the confideration of the general equation

reprefenting them.

The

general

equation of 21. by extermi-

nating the fecond term can be transformed into

the equation, a% *A
,
,

ah id

From which we have

Where

the

two values of z
figns, fo
abfeiffes

are always equal,

line

and
the

have contrary

that the

DF, on
of

which the
ordinates,

are taken,

muft
is

bifeel:

and

confequently,

a diameter

the conic fedtion.

And,

as this

has been de-

monftrated generally,
lines

in any

fituation

of the

PM,

it

follows that if any parallels, as

Mm,

340

A
is

Treatise
which can

of

Part III.

Mm, Mm,
there
rallels.

be drawn meeting a conic feclion *,

a line

DF

bifeel all thefe

pa-

And

cor.fequently if any

two

parallels,
line

Mm, Mm,

are bifecled in

and g, the
bifeel;

Gg

that bifects thefe two, will

lines parallel to
Ci

all

the other

them, terminated by the curve.

all

Which
There

is

a general property of

the conic

feclions."
is

one cafe which muft be excepted,

is
it

when
in

PM

parallel to

an afymptote, becaufe

in that cafe

meets with the conic fection only

one point.

2$. In the general equation of 21, if you z o, there will remain cx dx e=zo, fuppofe jy

by which the points

curve meets the

are determined

where the
and that

abfcifle
it

Suppofe

it

meets

AE. in B and D,
Then
fhall

AB =

A, and

AD == B.

A and

B
+.^-

be the two roots of the equation x z

+ x
x%

=.0 j and

therefore x

+Ax x+B=

+ lx + L C

but# + A = BP,and#+B = DP;

* Supply the
figure.

therefore

Chap.

2.

ALGEBRA,
BP x DP ='x z + x +
d
e
.

341

therefore

Now,

it is

manifeft from the nature of equations, that

if

meet the curve in and of the roots

PM
cx
z

M and m,
Pm
z

the rectangle

PM
+

mail be equal to

+ dx +

e the. laft
z

term of the equation

+ cx by + dx +e We have therefore PM x "Pm == cx + dx + e\ andBPxDP = * + y* + ^i fothatPMx

y
axy
-f-

Pm:
:

BPxDP
That
is

:: cx

+ dx + e:

+ -x+-::c

1.

is,

"

the rectangle of the ordinates

PM, Pm
Which
is

to the rectangle of the fegments of

ratio, c
is

the abfciffes, as, in a given

to i."

another general property of the lines

of the fecond order.

In a fimilar manner the analogous properties

of the

lines

of the higher orders are demon-

ftrated *.

36. There are

many
is

different

ways of deby mo-

fcribing the lines of the fecond order 9

tion.

The

following

Sir Jfaac Newton's,

Let the two points

and S be given, and

fame plane.

the flraight line

AE

in the

Let the

* See the Appendix.

*f

See Geomeiria Qrganica, Prop.

I..

given

342
given
points

^Treatise
angles

of

PartHL
the interfecti-

FCO, KSH,
as poles,

revolve about the

let

and S
fides

and

on of the

CF* SK, be

carried along the

ftraight line
fides
'order.

AE,

and the interferon of the

.

CO,

SFI, will defcribe a line of the fecond

Let the

fides

CF, SK

interfecl
in

each other in
:

Q,

and the

fides

CO, SH,

let

PM

and

QN
and

be perpendicular on CS.
;

Then draw PR,

QU; PT,QL fothatCUQj=CRP=FCG-,

SLQ^ STP - KSD. The angle RCP = CQU,

right ones with

two

the triangles

CUQ

RCP makes RCQjind QUC. So that and CRP will be fimilar.

fince
flrate

And

after

the

fame manner you may demon-

Chap.

2.

ALGEBRA.
SLQ, STF

343
are fimilar^

firate that the triangles

whence,

and
Suppofe
angle

CR ST
CS
to

PR PT

QU
QL
,

CU,
SL.
the
fine
;

= a, CA =
its

of the
angle

FCO

cofine as

to a
fin.

fin.

CAE

to cofin. as c to a 9

and

KSH
-,

to cofin.

as e to a.

Put

alfo

PM = y, CM =

QN = z.

ThenRM:PM::^:i. PR:PM::^+^:(/.
AN.-QN::*:^. So
that

RM = ^, CR(=CM

^RM)^^l-|/k-/^^
Likewife

*+ QU = z,/a d\
a
c

a nd

CU (= CA
And
it

AN NU) - -z
b

a
-,z. a
it

be-

ing

CR
:

PR

QU
:

CU,
2

follows
b

that -rZ.

i*

#j

jy\V -|- d 2

z>/tf

+d

z
:

+
ax

So that z

ay dc j Xy + d +
be

X dx

cX

In like manner you will find

ST=za

^-

(= AN AS NL)

= a b + "??***
Z
3

But
ic

344
it

<

Treatise
: :

of

Part IIL
is,

was
:

ST PT
:

QL
z

SL,
z
:

that

a
e

x
..

y
e

ys/a -\-e
-

%s/ a
:

->r e

a
b

bA

cxa%
ec

Whence QN

== 2J==

X c X ae ex

ay

ec^a-xy-^axxe-c-^-a^xc-e

And from
z
a
this

the equation of thefe

two values of

equation refults
ce

b X
-{-

\ x z +a r

xd+c~el
r

xy

+ ax-a
\y

-\-cdl
.
.

+ ae-kcxd 3
abc

+ dee
d
4- el

3
z

bcxe + dy\y
az )

-\x bexa -edl exd + aexde

x
az
)
-t

= o>

where
it

and y are only of two dimenfions s appears that the curve defcribed muft be 3
fince x

line

of the fecond order ^ or a conic fection, ac-

cording to what has been already demonftrated.

3 7. As the angles FCO, the poles C and S, if the angle

KSH revolve about CQS becomes

CPS
muft

equal to the liipplement of thefe given angles

to four right ones, then
vanifh, that
is,
:

the angle

the lines

CO

and

SH

muft be-

come
and

parallel

and the interfeclion P muft go

off to an infinite diftance.

And

the lines

CO

SH

become,

in

that cafe, parallel to one

of the afymptoies.
In order to determine
if this

may

be, defcribe

on CS an
in
it

arc of a circle that can

have inferibed
If
this

an angle equal to the fuppiement of the

angles

FCO, KSH,

to four

right

angles.

Chap.
this arc

2.

ALGEBRA.
line

345
#,

meet the
to either

AE

in

two points N,
as
it is

then

when Q^

the interferon of the fides

CF,
car-

SK

comes

of thefe points,

P
ried along the line

AE,

the point

to infinity, and the lines

rallel

P will go off SH, CO, become pa-

to each other

and to an afymptote of the

curve.
If that arc only touch the line

AE,
it,

the point
If the arc

will

go off

to infinity but once.

neither cut the line

AE

nor touch
In

the point
firfl

P cannot go

off to
is

infinity.

the

cafe

the conic fection

an hyperbola, in the fecond

ellipfe.

a parabola^ in the third an

The

afymptctes,

when

the curve has any, are

determined by the following conft ruction.

Draw

NT

conftituting

the angle

CNT

=p

SNA,
I

meeting

SC

in

then take SI
parts,

= CT,

and always towards oppcfite

and through

draw IP

be

to SH or CO, and I? will one afymptote of the curve. The other is

parallel

determined in like manner, by bringing Qjto

.

Z 4

And

34^

A
the
this

Treatise

of
in

Part III
the center,

And

two afymptotes meet

conftruction
is

conftituting there an angle

== NS.
it

From
when

cb bus, mat
line

the circular arc

CN;?S touches the

being then

AE,
line

the angle

SNA

= SCN,
-,

the

NT will
CT

become

parallel to

CS

and thereis,

fore

and SI become

infinite,

that

the

afymptote

IP going

off to infinity, the curve be-

comes

a parabola,

38.

There

is

another general method of de-

ferring the
ferves

lines of the fecond order^ that de-

our confederation.

we now ufe DQ, CN, SP, which we fuppofe

Inftead of angles

three rulers
to

revolve

about the poles D, C, S, and cut one another

always in three points N, Q^and

and carry-

ing any two of thefe interferons, as

and Q,

along the given flraight

interferon

lines

AE, BE,

the third

will defcribe a conic feclion.

Through

Chap.

2.

ALGEBRA.
the

347
F,
to

Through
P1VI,

points

D, P, Q, draw DF,

QR, parallel to AE, meeting CS in M, R; alfo through P draw PH parallel

BE

meeting

CS

in

H.

AB {= a PMH, AEB

Then putting PM = y, CM = x, CSz=a, CA = , SB = c, DF = &, AF = /, AE^J, BE = e,

h ,+ c)

=/j

fince the triangles

are fimilar,

therefore

PH

= -|,

MH = 4
;

SH

= * + =f

And

fince

CA AN:; CM: PM,...AN = -^; X

But,

and

fince

?B:Ba::SH:PH,...B(^=^_^
PQj QR
:
:

BE AE
:

QR
BR

cdy

dx-t fy

cf
,

and ...
:

=
::

SS
:

N QW AN DF RQ^-AN
that
is,
'

AF AR;

a-

dx-\-fy

tfJ

a-

dx-\-fy

^*

And

multiplying the extremes and

it
is,

means,

and ordering the terms,

bfxc-l-fxy* + cxld-kf-bdxl+f + kffxxy

+ badx l+fxy - adfkx x+dfk x x

may

}and

In which equation, the figns of fome terms

vary by varying the fituation of the poles

34 s
and
lines
;

^
but
sc

Treatise
and y not
it

of

PartllL more than

rifing to

two dimenfions,
ways
defcribes
cafes

appears that the point

al-

a conic fcclion.
the
for

Only

particular

conic

feclion

fome becomes a
in

ftraight line.

As

example, when
;

is

found

in the ftraight line

z the terms dfkx

adfkx

CS

for then

DF

vanifhing

vanifh, and the remain-

ing terms being divifible by y 9 the equation be-

comes,
bfx c-4-jxy

+ cld-bdxl+fxx
cafe,

-f-

badxl+fz=. o.
and fhews,

Which
that,
line.

is

a locus of the
this

firft order,

in

P mull
it

defcribe a ftraight

After the fame manner

point

appears that if the

lines

the interferon of the

AE, BE,

falls in

CS

then will
cafe

defcribe a ftraight line.

For

in

that

vanifhes, and the equation

becomes.

bx

clfxyfX c kx x = o.

$g. Thefe two defcriptions furnifh, each, a general method of " defcribing a line of the

fecond order through any five given points where-

of three are not

in the

fame ftraight

line."

Suppofe the

five

given points are C, S,

as

M, K,
to the

N
let

join

any three of them,

C, S, K, and

angles revolve about

and S equal
and

angles

KCS, KSC.

.Apply the interlection of

to the point Hfj
let

the legs

CK SK
?

firft

the

inter-

Chap.

2.

ALGEBRA.
to the remaining point

34$

interferon of the legs CO and SH be Q^*, fccondly apply the interferon of the fame legs

CK, SKj

M, and

iet

ths

interferon of the legs

a line joining

CO,

SH
it

be L.

Draw
line

and L, and
if

will

be the

AE

along which

you carry the interferon of

the interferon of the other

the legs

CO, SH,

legs will defcribe a conic fe&ion pafling

through

the five given points C, S,

It

M, K, N.
and S from the con-

muft
:

pafs through

ftru&ion

when
A,

the interferon of

CO,

SH
K.

comes

to

the curve will pafs through

And

35
N, M.

Treatise
to

of
it

PartlXI.

And when it comes

Q^and L,

paries

through

40.

From
S,

the fecond defcription

we have

this folution

of the fame problem.

Let C,

M, K
till

N be the five given points

\

draw
lines

lines

joining them

produce two of the

NC, MS,

they meet in D.

Let three

rulers revolve about

viz.

the
line

D, CP, SQ 5 DR. Let the 'interfe&ion of rulers CP, DR, be carried over the given
the three poles C, S,

MK,

SQ,

DR

be carried

and the interferon of the through the line

rulers

NK

and the point

the interfe&ion of the rulers

that

Chap.

2.

ALGEBRA.
C
and S,

351

that revolve about

will defcribe a conic

fe&ion that paffes through the five points C,

S,

M, K, N.

41. It is a remarkable property of the conic fcctions, that * if you afTume any number

of poles whatfoever, and make

about each of them,

rulers- revolve

and

all

the interferons
lines, that

but one, be carried along given right

one mail never defcribe a

fcdlion
;'"

line

above a conic

if,

inftead of rulers

you

fubftitute

given

angles which

you

move on

the

fame

poles, the curve defcribed will Hill be

no more

than a conic feclion.

By

carrying one of the interfedrions necefTary

in the defcription over a conic fee"lion, lines of

higher orders

may

be defcribed.

CHAP.

352

Treatise

of

Part III.

CHAR
Of
42."fir^HE
JL
of a

III.

the Conftruftion of cubic

biquadratic Equations.
roots of any equation

and

may
di-

be determined by the interferons

ftraight line with a curve

of the fame

mensions as the equation :" or,

" by

the inter-

ferons of any two curves whofe

index of the propofed equation.
3*

indices multi-

plied by each other give a product equal to the

Thus

the

roots

of a

biquadratic

equation

may

be determined by the interferons of two

,

conic fe&ions

for the equation

by which the

ordinates from the four points in which thefe

conic leclions

may

cut one another can be defour dimenfions

:

termined will
conic feclions
ner, as to

arife to

and the

may be affumed in fuch a manmake this equation coincide with any

:

propofed biquadratic

fo

that

the

ordinates

from

thefe

four interferons will be equal to

the roots of the propofed biquadratic.

If one of the interferons of the conic feclion
falls

upon the

axis,

then " one of the ordinates

thefe ordi*

vanishes,

and the equation by which

nates

are determined will then be of three di*

mentions only, or a cubic" to which any propofed cubic equation

may be accommodated.
So

Chap.
So

3.

ALGEBRA.

353

that the three remaining ordinates will be

the three roots of that propofed cubic.

43. Thofe conic feclions ought to be preferred for this purpofe that are moft eafiiy defcribed.
cles;

They muft

not however be both cir-

for their

interferons are only two, and

can ferve only for the refolution of quadratic

equations.

Yet the

circle
;

ought to be one,
and the parabola

as
is

being moft

eafiiy defcribed

commonly

afTumed for the other.

Their interferons are

determined in the following manner.

Let

354
Let
bola.
its

-rf

Treatise
common
its

of

Part III,

APE

be the

Apollonian para^

Take on

axis the line

AB

== half of

parameter*

Let

be any point in the plane

it

of the parabola, and from with any radius

bola in P.

a a center defcribe,

CP,

a circle meeting the para-

Let

FM, CD,

be perpendiculars
let

on

the axis in

M and D,
PM
in

and

CN,

parallel

to the axis, meet

N.

Then

will

always

Put CP = a> t AD = DC p% AM = a, PM y. Then CNg = x | c\\ NP? =jf$j and * That T^TT +yYd\ = ** -\-y + zdy + d = # x + 2cx +
9

(47 e r.) the parameter of the parabola

CP^Clty + NPf

c,

ft*
l

But, from the nature of the parabola, j* == &xv

and ^ z =r
lues

fubftituting therefore thefe va#,

z

for x*

and

it

will be,
c
z

y ^f + y 2dy + +
Or, multiplying by %

dz

-z o.

Which may

reprefent any biquadratic equation

-,

that wants the fecond term

fince

fuch values

may
this

be found for

a, b 9 r,

and d9 by comparing,

with any propofed biquadratic, as to

coincide.

make

them

And

then the ordinates from

the points P, P, P, P, on the axis will be equal

to the roots of that propofed biquadratic.

And
which-

* x

cr.c is

the difFeiences of
is

Ur

and

c indefinitely,

ever of the two

greateft.

this

Chap.
this

3.

ALGEBRA.
:

355
if

may

be done, though the parameter of the

that
is,

you have a parabola already made or given, by it alone you may refolve all biquadratic equations, and
parabola (viz. b) be given

you

will

only need to vary the center of your

its

circle

and

radius.

C
Ss

44. If the circle defcribed from the center pafs through the vertex A, then CP^z=CA^

CD^

AD^,

that

is,

az

-\-

z
\

and the

lad term of the biquadratic (c 2

variifh
;

+d a
z

will

therefore, dividing the reft

by y, there

arifes the cubic,

y** +. % bc

+ b X y + zdb
z

== o.

Let the cubic equation propofed to be refolved b e y** Compare the terms of r py

thefc

356
thefe

^Treatise
two equations, and you
and

of

PartllL
have

will
or,

ibc

-[-

2,

+ p> =~r

+ + idb =
z

r,

Tfr-"!^ 2 2
'

and

^= + 7T.

From which you have

this

conftruction of the cubic

y^^py +

= o,
is

by

means of any given parabola APE.

" From

the point

take in the axis (forward if the

y

equation has
tive) the line

p
=.

but backwards if p
;

poji-

BD = ~

then raife the perpen-

dicular
circle

DC

and from C,

defcribe

paffmg through the vertex A,

meeting

the parabola in P, Jo fhall the ordinate

PM
+
*"

he one of the roots of the cubic

= or
The
ordinates that
ft

y *+.py
3

and on the fame

fide

of

the axis with the center

are negative or afis

firmative, according as the laft term r

tive

nega-

or affirmative %
contrary
axis.

and thole ordinates have

that

always
fides

figns

Hand on

different

of the

The
it

roots are found of the

figns,

fame value, only they have contrary

r
is

when
which
elfe-

pofuive as when

is

negative

the fecond
,

term of the equation being wanting

agrees with

what has been demonftrated

where,
45.

In refolving numerical equations,

,

you
then

may

fuppofe the parameter b to be unit

AD

Chap,

3.

ALGEBRA.

357

AD = + ip, and DC = ir PM muft then be meafured

the parameter,

and the ordinate

on a

fcale

where
be

or

2AB

is

unit.

Or,

if it

to exprefs 10,

more convenient, the parameter may be fuppofed 100, &c. or any other number,

and
fcale

PM

will be

found by meafuring.
is

it

on a

where the parameter

10,

100, &c. or

that other number.

46.
in

"When

the circle meets the parabola

one point only befides the vertex, the equatireal

on has only one

imaginary."

root, and the other

two
falls.

Thus,

if

the equation has

fide

+ p,

or if

on the fame

of

as

A does,

the circle can

meet the parabola

in

two points only, whereof

as

is

one

and therefore the equation muft

;

have two imaginary roots

el fe where.

we demon (1 rated

If the circle touch the parabola, then

two

roots of the equation are equal.

It is alfo

obvious, that the equation muft ne;

cerTarily

have one real root

one other point,

becaufe, fince the

it

circle

meets the parabola in the vertex A,

it

rnaft

meet

in

at leaft,

befides

A.
9

47. Inftead of making the circle pafs thro the vertex A, you may fuppofe it to pafs through

fome other given point

in the

parabola, and that

interferon being given, the biquadratic foundfor determining the interferons, in 43,

may
Let

be reduced to a

cubic.

a 2

358
Lee
the

Treatise

of

PartllL
given

ordinate

belonging to that

inter feclion

be , then one of the values of

follows that the biquadratic
t
z

being g,

it

y**

^Z y *+ 2 db y + d +
by y

-a xb 2 =iO
z

will be divifible

g,

which

will

reduce

it

to a cubic that fhall have the fecond term. thus

And

we have

a
all

conftruction for cubic equatitheir terms.

ons that have

For example,
meter
is

let

us fuppofe that the paraat

AG,

and the ordinate

G
F

is

GF

meet-

ing the curve in F.

circle is

Suppofe now that the

;

always to pafs through

a*)

then fhall

CF? (=
c
z

= CH? + HF? = 7+T + T+aX = -f d* _ icb zdb + 2b\ and fubftituting

in

Chap.
in

3.

ALGEBRA.
43
this

359
0%
it

the equation of

value of

becomes
j+*

icb

ly l

*db z

2b* it*
-J

+ -#$
Where
to
r in
it

2C&

>

= o.

the laft term has a contrary fign

what

has in the third, and d a contrary fign

has in the fourth.

to

what
roots

it

This biquadratic has

its
;

FG,

or

b,

and being divided by y

=
C

for

one of
b,

there

arifes this cubic,

yi

_j-

hy

2ci?1

+ ib
having
all its

2db'' v I j if*

2f* >
?

O,

+ 2^
terms compleat.

J
If

had been

taken on the other fide of the axis, the fccond

term by z had been negative.

Let now any cubic equation be propofed to

z be refolved, as y l r=o. And, by qy py comparing it with the preceding, you will find

p=z

p-A
,

q-=.2b ZJr 2bc *

> whence^
L

.+c=zp r ^

-,
2p*

rz=.2b lJr2cb XJr2db'

Therefore, to conftruft the -prop fed cubic equation

+ py +
z

qy

=
r
to

o,

let

the parameter ofyour

parabola be equal

p, take,

on the axis

from
the

Aa

360
the vertex

A
A^

Treatise

of

Part III.

the line

AD = p

and

raije

the perpendicular
defcribe

DC

= + t>
F
7

an &

from C

circle

through

meeting the parabola

in

fo JJoall the ordinate

PM be

a root of the

equation.

If the equation propofed Js a literal equati-

on of
all the

this

form y z
vvili

+ py

+pqy

r w? o, having

terms of three dimenfions, then

only require

struction

AD =

this
-Jf

conanr^

48.

If

through any point

you fuppofe the parabola to pais F taken any where in the

call the ordinate
,*

parabola (vid. Fig. preced.) and

FG = e

Vp then c -j-

= a\
d] this

and the

general biquadratic

may have

form,

But

fince

FG ==

<?

is

the equation will be divifible by

one of the values of y, e> and the y

quotient

is
i

found to be
ey
z

this

cubic?

+ +

%bc"\ b
2

zdb*
zceb
rf*

+ +

>

^
Which

Chap.

3.

ALGEBRA,
1

361

Which compared withj Jrpy z + pqy

gives

p rz=io
1

FG (or e) z=p, AD (= c)

and

DC = ^ = *

And by

this

con-

ftru&ion the roots of a compleat cubic equation

may

be found by any parabola whatfoever.

49. It is eafy to fee from 43. how to conftruct the roots of a biquadratic by any parabola,
after

the fecond term

is

taken away.

But " the roots of a biquadratic may be determined by any parabola :" only they cannot be the ordinates on the axis, but " may be equal
to the perpendiculars on a line parallel to the
axis,

meeting the parabola in F,

CD

in

H, and

PM in L."
Let

FG

be an ordinate to the axis in

and

the reft remaining as before, let

FL=tf, PL=_y,
the

362

Treatise

of

Part III,

the parameter

= ,

CP = ^, FH = c,

CH = ^,

jy

= AM x h therefore + ^ = AG + FL x h = y + and eonfequently y + 2^ = fo. j- + # X # * But CN# -f NPj = CP^t, that -f if = a And fubftituting for x and * their
And
fince
z

FMq (= PL + HD]?)
-ley
z

is,

c\

values -

and

*"*"

^
;

you

will find

which
4

is

a complete biquadratic equation.

it

And

by comparing with

the equation

+ Py + %J
%

~"" b"~ry

J>

= o,
and

you

will find

HC (= # = a/^j + +

a
z
ii
:

Jg^,
which gives

CP (4)

a general conftruc-

tion for any fuch biquadratic equation by any

parabola whatfoever.
or
s,

If the figns
is

of p,

q,

rx

are different,

it

eafy to

make

the ne-

ceiTaiy

alterations in the confcruclion.

Ex.gr.

If

p
If

is

negative, then

FG

mufl be taken on the

other fide of the axis.

you fuppofe

the circle to pafs through F,

the equation will

become

a cubic Jiaving

all

its
:

terms

Chap.
terms
:

3.

ALGEBR A.
laft

363
z

the

term
z

d
-f-

+d
z

x U
It

vanish-

ing, becaufe then c

this

d e ah

will

have

form,

+
and then
cc

*->

the conirruction will give the roots

of a compleat cubic equation."

have fufRciently fhewed, how the 50. roots of cubic and biquadratic equations may be
conftrucled by the parabola and circle

We

now fhew how

other conic fections

we dial may be de;

termined by whofe interferons the fame roots

may
z

be difcovered.

Let the equation propofed be j 4-* -f- bpy z -fo ; and let us fuppofe, that, b r b qy z i. bx~y \ then fhall we have by fubfti-

=
b

tution of b x z for
2
.
.

, z

-\

y\ and
bq

dividing by bp 9
b*r

x-

+
,

y
p

= o,

which has
(in

its
.

locus
laft)

an

r effipfe.

Then by

fubftituting
all

this

bx for

y\

and multiplying

the

terms by y, you find,

x 43 a parabola.
.

px

+ oj br = o,
Then,
adding to

an equation to
this

bx =
4
.

equation

o,

you

will have,

+ /- J__ Z
circle.

>x

gg

=
r

o,

an equa-

tion to a

The

364
The
%

Treatise

of

PartllL
bpy*

roots of the equation

b r zz o

y^ -f
-,

+ b qy
z

may be determined by
loci

the interfec-

tion of any two of thefe

as for

example,

by the interferons of the

f the equation y
and of the
%"
circle
z

-\

ellipfe

that

is

the locus

-J- ~jf

= ,

which

is

the locus of
0,

+y

_"? f # + $}

if =

from which we

deduce

this conftru&ion.

La


Chap.
Let
3.

ALGEBRA.
be
the
let

365
equal to

AB
-\

axis of

an

ellipfis,

,J br
raife

be the center of the ellipfe^

and the

axis to the parameter as

to

to b.

At
it

G9

a perpendicular

the axis,

and on

take

GD =

and on the

other Jide in the perpendicular

continued take

GK

= \q

x
P
:

Let

DE
ib

and
9

KC

be parallel to the axis

center*

take

KC c=

lp

and from C as a
defcribe

with the radius

ellipfe in

V DCq-\- br
P, and the

circle

meeting the

ordinate

PM,

on the line

DE,

Jhall be one of the

roots of the propofed equation.

PM (=x) produced meet AB in R, and KC in N and calling DM = #, then CFq = nph-nc?, that iq + i^~ ipb+ip ip x\ -]~y-\-i and therefore, + brc=:ib + * + qy ~T ^ >x brzzo, thecquaLet
;

is >

';

r q]

i. jy

tion to the circle,

which was to be conftructed.

:

A nd
fore

fince

P%
:

GBq
ap

- GRq
x%
:
:

::b:p,
b
:

there-

y J

+ T ~! m
fJ
P

br

p r

and con-

fequently,
y which ; P P the equation that was to be conftrucled.

xz

-4-

2= o

is

Now

that

their interferons will

give the

roots required, appears thus.

For

366

^Treatise
firfl

of

Part III.

For x 1 in the lue you deduce

br

equation fubftitute the vait

for

from the fecond, viz,

will arife
.

-j-y

qy 9

and there
b

bl
9

p
y*
;

that

is,

~ == x
y

and x z

= j^

which

fubftitut-

ed for x z and x

in the firft equation, gives

y:
is,

+ f + P1^ Xj+?)-^ = Oi == ^7 j *
4

that

-}- //" -[if

b>r

o.

And

you
bq

fubftitute

them

in

the

fecond

equation, there will arife

w+
.

y'+y
z
2

. z

+jy
%

~ ==
b*r

o,

that

is,

bpy

b qy
it

b =
r

p
o, the very

j 4*
-

J+

'

fame

as before

and thus
^4*

appears that the roots of the equation

__[_

fyyi

pqy __

lir

are tne or dinates

ellipfe,

that are
are

common

to the circle and

or that

drawn from

their intellection.

End

of the

Third Part,

APPEN

db^gfedMbe^db^^^^o^^^^^^^gfegfe^

appendix:
D E

Linearum Geometricarum
Proprietatibus generalibus

TRACTATUS.
^^^3?5^S5^53?52?53?5S55^5o3p^5^^^S^5Sp^5^5^

D E

LlNEARUM GEOMETRICARUM
Proprietatibus generalibus.

^j(3WD{( E lineis fecundi ordinis, five fetfonibus co&& j) W^ nicis, fcripferunt ubcrrime geometrae ve-

Sf^^T^ TKsl sa.7f\

teres

& recentiores

de

figuris quae

ad fu-

p er j ores

Hnearum ordines referuntur pauca

&

exilia

tantum ante

Newtonum

tradiderunt.

Vir

illuftriflimus,
tertii

in Tradlatu

de Enumerations Linearum
diu jacuifTet, ex-

Or dints, doctrinam hanc, cum

dignamque
Expofitis
efle

citavit,

in

qua elaborarent geometris

proprietati-

oftendit.

enim harum linearum

affines ut velut

bus generalibus, quae vulgatis fedtionum conicarum af-

fe&ionibus funt adeo

ad eandem nor-

mam
ut
verfis

compofitae videantur, alios fuo exemplo impulit

five

analogiam hanc
intercedit

fimilitudinem quae tarn di-

figurarum generibus bene cognitam

&

fatis

firme animo conceptam atque comprehenfarn

habere ftuderent.

In qua illuftranda

&

ulterius inda;

ganda curam operamque merito pefuerunt

(it

cum

nihil

omnium qua

in difciplinis pure mathematicis tra-

clantur quod pulchrius dicatur, aut ad

veftigandi

cupidum obleclandum

aptius,

animum veri inquam rerum

tarn diverfarum confenfus five harmonia, ipfiufque doclxinae priori

compofitio
convenit,

&

nexus admirabilis, quo poflerius

fequitur
fuperiori

quod

refpondet,

quaeque

37 JD^LinearumGeometricarum
quaeque fimpliciora funt ad magis ardua viam conftanter
aperiunt.

Linearum

tertij

ordinis proprietates generales a

New-

iono traditse parallelarum Tegmenta

&

afymptotos pie-

rseque fpeclant.
verfi generis

Alias

harum

affetiones

quafdam di-

breviter indicavimus in traclatu de fluxi-

onibus nupcr edito> Art. 324,

Cotejius

&

401.

Ceieberrimus

pulcherrimam olim detexit linearum geometri-

carum proprietatem, hucufque ineditam, quam abfquaf demon ftratione nobis communicavit vir Reverendus D.
Robertus Smith, Collegii
S. S. Trinitatis

apud Cantabrigi-

enfes Prsefeclus, doctrina operibufque fuis pariter ac fide

&

fludio in amicos clarus.

De

his

meditantibus nobis

quoque fe obtulerunt theoremata generalia; quse cum ad arduam hanc geometrise partem augendam & illualia

itrandam conducere viderentur,

ipfa quafi in

fafciculum

congerenda

&

una

ferie

breviter

exponenda

&

demon-

ftranda putavimus.

E C T

I.

De

1.

Lineis Geometrich in genere.

T
JL/

Ineas fecundi ordinis feciione folidi geometric?,-

coni

fcilicet,

definiuntur,

unde earum proprie-

tates per

vulgarem geometriam optime derivantur.

Ve-

rum

diverfa eft ratio figurarum

qus

ad fuperiores line-

arum
rales

ordines referuntur.

Ad

has deflniendas,

earumque
Reprae-

proprietates eruendas, adhibendae funt sequationes gene-

co-ordinatarum relationem exprimentes.

fentet

abfciflfam
b.

denotentque a,

<f

AP, y ordinatam PM figurae FAIR, dy e3 &. coeffirientes qaafcunque

a

inva-

Propriet&tibus generalihiis.
.

371

Invariabiles

&

dato angulo

APM
quae

fl

relatio co-ordina^

tarum x

hy

definiatur aequatione quae, praeter ipfas co-

ordinates, folas involvat coefficientes invariabiles, linea

FMH
y

geometrica appellatur
linea algebraica,
lineas

quidem aucloribus
ipfius

quibufdam

aliis

linea rationalis dicitur.

Ordo autem
Vel a

pendet ab indice altiffimo

x vel

in terminis sequationis a fra&ionibus

& furdis liberate,

fumma
2,

fumma

indicis utriufque in

termino ubi haec

xy,

prodit

maxima.

Termini enim #%
;

ordinem

pariter referuntur

y ad fecundum 1 termini x 3 ^ x*y y xy' ^ y 3 ad

tertium. Itaque aequatiojf =z ax -f- b, five y defignat lineam five eft primi ordinis

ax
b

o,

&

locum primi
Fig. 2,

ordinis* quae

quidem Temper

rec~ta eft.

Sumatur enim
fit

in ordinata

PM

recla

PN

ita ut

PN

ad

AP

ut

ad unitatem; conftituatur
aequalis ipfl -j- b%

AD

parallela ordinatae
redtas

&

ducla

DM parallela

locus cui aequatio propofita refpondebit.

PM AN erit Nam PM rz:

b.
Gl

PN
AP

+ NM =2
fit

(a

X AP

+ AD)
vel

aequatio

formae y

z=zax~b

ax + Quod ax -\-b, recla y

AD,
j

vel

PN, fumenda

eft 'ad

alteram partem abfciflse

fitus contrariis coeffici-

contrarius

enim reclarum

entium

fignis refpondet.

Si valores affirmativi ipfius

defignent re&as ad dextram duclas a principio abfciflae

valores negativi denotabunt rectas ab

x A,

eodem

principio ad

finiftram

du&as

&

fimiliter

fi

valores atfirmativi ipfius

conftitutas, ne-

y ordinatas repraefentent fupra abfciflam

partes ductas.

gativi defignabunt ordinatas infra abfciflam ad oppofitas

iEquatio generalis ad lineam fecundi ordinis

fermse
yy

eft

hujus

axy -fw^o
**-

by

dx
*

+
Bb

<\yi

jD<?LinearumGeometricarum
n
k

&

sequatio generalis ad lineas tertii ordinis eft y 3

ax-\-b

Xy
Et

-\-

cxx

dx + eXy

fx*'-\-

gx

hx

-(-

o.

fimiiibus aequationibus definiuntur lineae geo-

metries fuperiorum ordinum.

2.

Linea geometrica occurrere poteft

recuse in tot

punctis quot funt imitates in

lineas

numero

qui sequationis vel

pluribus.

ordinem defignat,

&

nunquam in

Oc-

curfus curvse

&

abfeiflae

AP definiuntur
Linea
tertii

ponendo yz=z0 9

quo

in cafu reftat

tantum ultimus aequationis terminus

ordinis ex. gr.

quern y non ingreditur.

eurrit abfcifTae

AP cum fx>gx z -^-bx i~o,

oo
cur-

cujus

aequationis

fi

tres radices fint reales abfciiTa fecabit

vam

in tribus punctis.

Similiter in aequatione general!

eft

cujufcunque ordinis index altiffimus abfciiTa x aequalis

numero
jor,

qui linese ordinem defignat, fed

is eft

nunquam maautem aequarealis,

adeoque

numerus maximus occurfuum curvse

cum
que

abfciiTa vel alia quavis recTa.

fit

Cum
femper

tionis cubicae unica faltem radix

idem-

conftet de aequatione quavis quinti aut imparis cu-

jufvis ordinis

(quoniam radix quaevis irnaginaria aliam

neceiTario femper habet comitem), fequitur lineam tertii

aut imparis cujufcunque ordinis reclam quamvis afymptoto

non parallelam in eodem piano dudtam in uno faltem puncto neceiTario fecare. Si vero recla fit afympLinea
igitur imparis cujuf-

toto parallela, in hoc cafu vulgo dicitur curvae occurrere ad diftantiam infinitam.

cunque ordinis duo faltem habet crura

gredientia.
jufvis ordinis radices
riae,

in infinitum pro-

iEquationis autem quadratics vel paris cu-

omnes nonnunquam
nullibi occurrat.

fiunt imagina-

adeoque

fieri

poteft ut recla in piano linea? paris

ordinis duela

eidem

Ss

Pfoprietatibas generalibus*

dinis
3.

27%

iEquatio fecundi aut fuperioris cujufcunque ortot fimplicibus, a furdis

quandoque componitur ex
liberatis,

& fractis

in fe

mutuo

duclis quot funt ipiius

;

aequationis

propofitae
eft

dimenfiones

quo

in cafu figura

FMH non
Similiter
fi

curvilinea fed conflatur ex totidem reclis,

quae per fimplices has aequationes definiuntur ut in Art.\,

aequatio cubica

componatur ex aequationibus

duabus in

fe

mutuo

duclis,

quarum

altera

fit

quadratica

altera fimplex, locus

non

erit linea tertii ordinis

pro-

prie fie dita,

fed feclio conica

lineis

cum

recra adjunda.

Proprietates

autem quae de

geometricis fuperiorum
affirmandae funt

ordinum

generaliter
lineis

demonftrantur,

quoque de

inferiorum ordinum,

harum
pleant

ordines defignantes fimul fumpti

modo numeri numerum comlineae

qui

ordinem

ditae

fuperioris

denotat.

Quae de

lineis tertii ordinis {ex. gr.) generaliter

ftrantur affirmanda

quoque funt de

tribus reclis in

demoneodem

piano duclis, vel de fetione conica

recta fimul in

eodem piano
cui

defcriptis.

cum unica quavis Ex altera parte,

vix ullaaflignari poteft proprietas lineae ordinis inferioris

fatis generalis

non refpondeat affe&io aliqua linearum

ordinum fuperiorum.
eft cujufvis diligentiae.

Has autem ex

ill Is

derivare

non

Pendet baec dodrina magna ex

parte a proprietatibus aequationum generalium, quas hie

memorare tantum convenit.

4. In aequatione quacunque coefficiens fecundi termini aequalis eft exceflui quo fumma radicum affirmati-

varum

fuperat

fummam
eft

negativarum

&

ft

defit hie ter-

minus, indicio

fummas

radicum aflirmativarum

&
ge-

negativarum, vel fummas ordinatarum ad diverfas partes

abfcifTae

conftitutarum, aequales

effe.

Sit aequatio

neralis

ad lineam ordinis w,

Bb

f*

374 -D^LinearumGeometricarum
y*

ax\.bY> y
u
-\-

n ~~ T

+ cxx dx -f ^x/-2 &V. = 0,

\

fupponatur u.-zzy

pro j> fubftituatur ipfius va-

lor

& in
n ~~ 1

aequatione transformata deerit fe-

cundus terminus u

ut ex calculo, vel ex do&rina*

facile patet
:

aequationum paffim tradita

&

hinc quoipfius

que

conftat,
fit

quod per hypotheiln valor quifque

u minor

valore correfpondente ipfius y differentia

unde fequitur

fummam valorum
a

ipfius

u (quoipfius

rum numerus
v (quae

eft n) deficere
eft

fumma valorum

fumma

ax 4-

1?)

differentia

Xnz=ax

-\-bi

adeoque priorem

fummam

evanefcere

&

fecun-

dum terminum
affirmativos

deeffe in aequatione

qua u

definitur, vel

&

negatives valores ipfius u aequales fumSi itaque

mas

conficere.
reclae

fumatur

PQ^=

'

Ht

fit

QMr=:,
terminatae

ex utraque parte puncti Q_ad curvam

eandem confident fummam.

Locus autem

Z- 3-

puncti Q^eft reta

BD qua? abfciffam ultra principium A

B
ita ut

produclam
ipfi

fecat in

AB == & ordinatam AD
,

PM parallelam in D ita ut AD = x h enim haec re<la ordinate PM occurrat in puncto Q^, erit PQ_ad PB (feu + x) ut AD ad AB vel a ad n adeofit
;

fi

que

PQjn

ut oportebat.

Atque hinc conftat

ret.am

femper duci poffe quae parallelas quafvis lineae

geometrical occurrentes in tot pun&is quot funt figurae

dimenhones

ita fecabit ut

fumma fegmentorum

cujufvis

paral-

Prcprietatihus^ general!bus.
parallels ex

27

una

fecantis parte ad

curvam termlnatorum

femper

aequalis

fit

tera fecantis parte.

fummae fegmentorum ejufdem ex alManifeftum autera eft reelam quae

duas quafvis parallelas hac ratione fecat ipfam neceffario

efte quae fimiliter alias

omnes

parallelas fecabit.

Atque

hinc patet Veritas theorematisA^w/WtfK/,quocontinetur

proprietas linearum geometricarum generalis, notiffimae

feclionum conicarum proprietati analoga.

recta quae duas quafvis parallelas ad

In his enim

fe&ionem termi-

nates bifecat diameter eft,

parallelas ad

&

bifecat alias

omnes

hifce

fectionem terminatas.

Et

fimiliter recla

quae duas quafvis parallelas linear geometries occurrentes in tot punclis quot ipfa eft

dimenfionum

ita fecat

ut

fumma partium ex uno

fecantis latere confiftentium

fit

&

ad curvam terminatarum aequalis

fumms

partium

ejufdem parallels ex altero fecantis latere confiftentium ad curvam terminatarum, eodem

vis reclas his parallelas.

modo

fecabit alias quaf-

5.

In squatione quavis terminus ultimus,

eft

five

is

quern radix y non ingreditur, aequalis

fa&o ex

radici-

bus omnibus in fe mutuo duclis ; unde ad aliam ducimur non minus generalem linearum geometricarum proprietatem.

Occurrat recta
eritque

Sc

/*,

PM lines tertii ordinis in M, PM X ?m X ?p =zfx^ gx* -\-hx-r- k.

curvam
in tribus punclis I,

Fig.

Secet abfcifta

AP

K,

L &
;

AT,

gx* o, quo in cafu aequatio generalis dat fx 3 y determinandis, ut in Art. 2. hx~k-zz o pro his valoribus hx , tt> gx% ~\ 7 k> expofuimus. /Lquatiorus igitur*

AK, AL

erunt valores abfcifls x, pofita ordinata

....

tres radices funt

AI,

AK, AL;

adeoque haec aequatio

componitur ex tribus x

AI,
7

x ~-

AK,

AL

in
fe

Bb

376
fe

D^LinearumGeometricarum
ductis
;

~ - =2 *~AIx^AKx*--- AL = AP AIxAP-AK X AP AL = IPxKPxLP=yX PMxPffl

mutuo
eitque

#3

-f-

X P^

Faclum igitur ex ordinatis PM, Pw, P,v, ad punclum P & curvam terminatis eft ad factum ex fegmentis IP, KP, LP, recipe AP, eodem punelo & curva terminatis in ratione invariabili coefficients
unitatem.
Simili
reclae

ad

ratione demonftratur,

dato angulo
cujuf-

APM,

fi

AP, PM, lineam geometricam

pun&is quot
ipfa eft

vis ordinis fecent in tot

dimenfio-

num, fore femper factum ex fegmentis prioris ad puntum P & curvam terminatis ad factum ex fegmentis pofterioris eodem punclo & curva terminatis in ratione
invariabili,

tono 9

6.

In articulo praecedente fuppofuimus,

cum New-

reclam
I,

punclis

AP K, L

lineam

tertii

ordinis fecare in tribus

verum

ut theorema egregium red-

datur generalius, fupponamus abfciflam

AP

in

unico

tantum nuncio curvam

Fig. 4.

punclum A. Quoniam igitur evanefcente y evanefcat quoque x, ultimus squationis terminus, in hoc cafu, eritfx 3 g# 2 -h
fecare
;

fitque

id

(ii

fumatur

Aa

verfus

aequalis

A, &

ad punclum a

erigatur

perpendicularis
%

ab =z
Z
-

^ ^:/X
J
}

AP X a fm X P^,

+
fit

ab

=/XA?X P

unde cum

PM X

aecrualis

ultimo termino/v 3 **g# z

+ hx,
ut

Proprietatibus generalibus.
1st

377
unita-

in

articulo

praecedente

erit

PM

x Ym x P^ ad

AP

x bP z

in ratione conftante coefficients

f ad

tem.

Valor autem

reclse perpendicularis ab eft

Temper

realis quoties refta

AP

curvam

in

unico punc"to fecat

in hoc
-j-

enim

2 cafu radices aequationis quadraticae^*

gx
orfoli-

h funt neceflario imaginariae, adeoque

9

quam gg
dinis,

&

quantitas *J \fh

gg
A

\fh major

realis.

Cum
9

igitur

recta quaevis in unico punto

eft

fecat

lineam

tertii

folidum fub ordinatis

PM, Pm P^
Juncta

ad

dum

fub abfcifTa

AP &

quadrato diftantiae punfti

punclo dato b

in ratiorte conftanti.

Ab

eft

ad

Aa
g9

five

radius ad

cofinum anguli

MP,

ut s/\fh ad

&

Ab

= ,/

-z.

Idem vero punclum

b Temper con-

venit eidem rectae

abfcifTa

AP,

qualifcunque

fit

angulus qui

&

ordinata continetur.

fit
fi

7.

Sit figura feclio conica,

cujus aequatio generalis

-\- e -=z.

yy

ax
AP
4^

aequationis cxx

dx

xy

-J-

cxx dx

O vX fupra

5c Fig. 5

-{- e

=. o radices

fint irmginariae,

recla

fectioni

non

occurret.

In hoc autem cafu

quantitas

femper fuperat ipfam dd\

unde

cum

fit

dxez=.cXx~- ab ha =
cxx
2C
Sc erigatur

4-

^C
%

(fi
K

fumatur

perpendicularis abfciffae in a

s/ AC
ita ut

dd

'

ab =z

'

\c

= ^x a? + ab
z

==f xflP^fitb?*

que

PM x?m
unitatem.

cxx >dx-\-e>

erit

PM X ?m ad

ut c ad
fi

Itaque in fectione quavis conica

recta AP fe&ioni non occurrat, erit, dato angulo APM, retangulum contentum fub reclis ad punetum B confiftentibus & ad curvam terminatis ad quadra-

b 4

turn

37^ I>LinearumGeometricarum
turn diftantise punc"ti

a puncto dato* in ratione con^

flanti, quae in circuJo eft ratio sequalitatis.

Manifeftum

autem
ordinis

eft

eandem methodum adhiberi

abfciiTa fecat in

pofte lineaequarti

quam

duobus tantum punclis,

fecat

vel lineas ordinis cujufcunque

puniQtis binartp paucicribus

quam abfcifta numero qui figurae

in

ordineni

defignat.

8. Hifce praemiiHs, progredimur ad linearum geor metricarum proprietates minus obvias exponendas eo-

dem

fere

ordine quo

fe

nobis obtulerunt.

Utebamur
petito,

autem iemmate fequenti ex fluxionum doctrina

quodque

in

traclatus de
;

hifce

nuper

editi Art.

717,

demqnftravimus

harum tamen

aliquas per algebrarrt

yulgarem demonftrari pofle poftea obfervavimus.

Lemma.
entibus, ut

Si quantitatibus #, y, 2, #,

&c. fimul
fit

flu-

& quantitatibus
erit

X, Y, Z, V, &c.

factum

ex prioribus ad factum ex pofterioribus in ratione conftanti

quacunque, ^ i

-L ~ x y
'

~\
'

-4-
s
'

4-

&c.

*?

=r + +_+_ + &c.
tia,

Porro, brevitatis gra-

quantitates appellamus fibi

in fe

mutuo

reciprocal quafie

rum

mutuo duSarum factum

effe ipfius #,

eft unitas,

dj-a

cimus redprocam

&

ipfius y.

9. Theor. I. Occurrat recla quavis per puntlum datum ducla lima geometrica cujufcunque ordinis in tot
punclis quot tpfa
eft

dimenjionum

recla figuram in his

punclis contingent^ ahfcindant ab alia recla pofitione data

per

idem

punclum datum duel a fegmenta

;

tot idem

hoc

punBo ierminata

Is horum fegmentorum reciproca ean-

dem

Proprietatibus generalibus*

379
con-

dem femper
trarias

co?ificieni

fummam^ modo fegmenta ad

partes

pun&i

dati fita contrariis fignh affici-

antur.
Sit
P.

pun&um

datum,

PA &

Pa

refee quaevts duse Fig. 6.

ex P du&se quarum utraque curvam fecat A, B, C, &c. et a, b, c, &c. quot ipfa

in tot
eft

puncKs

dimenfio-

num.
#k 9
bl,

Abfcindant tangentes
cm, &c. a

duaa fegmenta
dIcoforc

AK, BL, CM, &c. et EP per pun&um datum P PK, PL, PM, &c. et Vk, P/, ?m &c,
re&a
9

TK + PT

+ Hl + &c =
-

Tk

+ 7 +

|-

&c. atque hanc

fummam manere

Temper ean-

dem manente pun&o P

&

reta

PE

pofitione data.

rallelis deferri,

Supponamus enim re&as ABC, abc motibus fibi paita ut earum occurfus P progrediatur in reta PE pofitione data ; cumque fit femper AP x BP

X CP x &c.
xionem

ad

aPx^Px
AP
BP,

c? in ratione conftanti per

ipfius

Art. 5. reprsefentet
retae

fluxionem

AP, BP

flu-

&

CP, EP, &c.

fiuxiones reta-

rum CP, EP,

bolorum

Szc.

refpe&ivas, ut vitetur inutilis fym-

multiplicatio, eritque (per Art. 8.)

AP -

-j-

BP -

+&c. = _ +
AP
eft

--

+ _ + *<:.
AP
EP
efie

Verumcum
notiffi-

rec"ia

motu

fibi

femper parallel deferatur,

ad

mum
nem

AP
EP

fluxionem re&ae
ut

EP

fluxio-

reclae

EP
'

AP

ad fubtangentem

PK, adeoque

AP AP
_

EP
5

PK
P

bimiIltCr

IF - PP
c? __

BP

EP
P/

*- p*

** c?

EP Ym

'

PM a? EP EP URde FK + PL
,
,

CP CP

EP

*P

380

ZVLinearumGiometricarum

+ "n:+FM + &c = H + F7 + pb+

-

&c

Haec
,
/,

ita fe habent quoties pun&a K, L, M, &c. -et w, &c. funt omnia ad eafdem partes punc~H P,

adeoque fluxiones rec~tarum AP, BP, CP, &c. *P, P,

cP, &c.

omnes ejufdem
puncla quaevis

figni.

Si

vero, caeteris

ma-

Fig. 7. nentibus,

et

cadant ad contrarias

partes

pun&i P,

turn crefcentibus reliquis ordinatis

AP,
rP,

BP, &c. earumque fluxiones pro

neceiTario

minuuntur ordinatae

CP &

fubdititiis feu negativis

funt

adeoque in hoc cafu

+
in

habendae
,

&c. =j
hifce

L
'

Yk

~-, r/
r/

&c.

&

generaliter

fummis

coiligendis, termini iifdem vel contrariis fignis aiEci-

endi funt, prout fegmenta cadunt ad eafdem vei ad

contrarias partes punCti dati P,

E,

10. Si recta

PE
ipfa

occurrat curvas in tot punclis

eft

D,
-

I,

&c. quot

dimenfionum, fumma

-j-

r -j~~

+
:

&c.

quam conftantem
aequalis
erit

feii

invariatam

manere oftendimus,

fummae

feu aggregate

FD

PE*

pT + & c
PE
;

e>

fummae reciprocarum
data?

fegmentis

reclas

pofitione
in qua,
fi

puncto dato

P
fit

& curva terminatis

ad
eft.

fegmentum quodvis

alteras partes

puncti P, hujus recrproca fubducenda

11.

Proprietatibus

generalibm.

381

libi

11.

Si figura

fit

fedYio conica, cui rela

PE

nul- Fig. 8.

occurrat, inveniatur

punctum

gatur P, huic ducatur ad retos angulos bd redfom

PE

fecans in d>

eritque

r +
ibY
5

112
==
'

b ut in Art. 7. jun-

Eft enim

PA X PB ad ,A?
.

bP z in ratione conftanti, adeoque (per

Art. 8.) -Tp

B? "bF

" .

= 7F

unde

(<l

uoniam
-

A h AP

eft

a.

ad

EP

ut

AP

ad

PK, BP

ad

EP

ut

BP

ad

PL,

&

b?

adEPut*Pad^~+^^
12.
Similiter
fi

||.

recta

EP

occurrat lineae

tertii

or- Fig. 9.

dinis in unico

puncto D, inveniatur punctum b ut in

b\r

Art. 6. reta bd perpendicularis in jun&arn

refta;

occurrat
eft

EP

in d,

&

quoniam

AP

BP x CP
PK.

ad

DP
"*"

x bP z
1

(ibid.) erit in ratione conftanti v

PM

PD

12 +
eae

4- -r-~
Px,
flt

Si

iV*

aUtem P ^ P er P endicularis
2

in

reclam EP, evanefcet

=r-=.

r#
Fig. 10.

13. Afymptoti linearum geometricarum ex data

plaga crurum infinitorum per banc propofitionem de-

terminantur

enim

confiderari poflunt

tanquam tan-

gentes cruris in infinitum product.

Recla

PA

afymv

ptoto parallels curvae occurrat in punctis A, B, &c. recla

autem
recla

PE curvam
ita ut

fecet in
=

D, E,

I,

&c. fiimatur in hac

PM

fit

aequalis exceilui

quo fumma
: -

Yd ^ TE"^n
*Zy~

&c

fLi

P erat

fumifla

'

+
:ro
les

&c.

&

afympcotos tranubit per

M;

382 jD<?LinearumGeometricarum
aequales fmt has

fummae, crus curvae parabolicum

erir3

afymptoto abeunte in infinitum.

14.

Ad

curvaturam linearum geometricarum unico

Fig 11. theoremate generali definiendam,

occurrant reda
fecet
recta.

PR

in

CDR circulus cui D & R, & reda PC in C & N"

fit

tan gens

CM

re&am

PD

in

M,

atque manente
deferri

DR,

fupponamus rectam

PCN
-

motu

fibi

femper

parallelo

donee coincidant puncra P, D, C,

&

quaeratur ultimus valor differentiae

-^=-. In recta

PN
ipfi

fumatur punSurn quodvis

;

q,

occurrat qv parallela

tangent!

in

CM reclae DR in v ducatur DQ_ parallela PN, & QV (parallela retae circulum contingent! D) fecet DR in V. Erit itaque -rrr ttft == I'M PD
'

^(quoniamDMxMR^CMO^p^p^^ _ = x MR X -^A ?M X MR X MD (cum MR FM i^

<v*
-j-

_
.

X MD,
Vv*-

feu

CM%
PM

fit

ad

PM
___ ~~~

ut

jw* ad
q<v

?v % )

q<V*X

X MR x PM + qv- X

i:'M a

P<z/

cujus ultimus valor s evanefcente

<ro

X MR + y* X PM* PM & coincidentibus

Z

& ?v cum QV & DV,

eft

QV

Atque idem
fi

eft

valor ultimus differentiae r-r

FM

-PD
D

D & C fint
cum
circulo

in

arcu lineas cujufvis ejufdem curvaturae

CDR.
Fig, 12.

*5- Theor.
trica:

II.

#
tot

punglo qiiovh

/iTtap

geome-

ducaniur du<z queevis reel a

utraque earn fecet in

punBis

DE, DA, quarum D, I, E, &c. &f D, A,

>

Bj &c, quot

ipfa &fl dimenfionum

abfeindant tangentes

AK

Proprietatibus generalibus.

383
&c.

AK, BL,

&c. a rt&a
Q^fcf V,

DE

fegmenta

DK, DL,

Occurrat recla qucevis,

QV
ra!?*

tangsnti

DT
2
ztf

parallela ipjls

DA
I
3

fcf

DE m

>?*

QV
zto

^ DV* /
771

fumatur in

DE

DR

-j- aquatis fit

excejfui

fumma
"I"

7 +

-frr

+ ^. fupra fuminam jr^

fupra chordam

"f"

tvT

^*

fcriptus

reclam

^ circulus DT contingens
cum
in

DR

*&-

erit circulus cfculatorlus^

jive ejufdem curvature

linea geometrica propofita^

ad

punclum D.
Oftendimus enim

fummam
-}- -pT

^ + ^S + fci l- + 'S
in Art. praecedente invenimus valodifferentiae

Art. 10.

(Fig. 6.) generaliter

+ &c. &
D &

rem ultimum
punftis P,

r
-rw*

-,

coincidentibus

C,

efle

DR

== ^Tr & circulus

ejufdem curvaturae
retae

cum

linea geometrica ad

punchim
r--

DE

occurrat in R.

Unde

fequitur fore

T3E+-5T+
procam
ipfi

&c

_ Eir-IlL- &c
effc

fivereci -

x DR
fummam

aequalem

excefliii

quo fum-

ma

reciprocarum fegmentis punclo

D &

curva termi-

natis fuperat

reciprocarum fegmentis eodem

pun<Sto
ties

&

tangentibus

AK, BL,

5cc.

terminatis.

Quo-

autem excefTus hie evadit negativus, chorda

eft

DR
ad-

fumenda
hibenda

ad alteras partes puncli

D, femperque

eft

regula fuperius defcripta pro fignis terminoSi reta

rum

dignofcendis,

DA bifecet angulum EDT

recta
,

384 jD^LinearumGeometricarum recla DE & tangente DT* contentum, theorema fie Hoc enim in cafu QV = DV, mz=zi f paulo fimplicius.
.&
t=--- sequalis

exceffui

quo

+ +
-

&c. fupe-

16.
nerate

Ex eodem principio confequitur theorema gequo determinatur variatio curvaturae vel menfura

anguli contaclus curva

&

circulo ofculatorio contend*

;

in linea quavis geometrica

pra^mittenda tamen eft ex-

plicatio brevis variationis curvaturae,

cum

hasc

non

fatis

dilucide apud auctores defcripta

fit.

Linea quasvis curva

a tangente fle&itur per curvaturam fuam, cujus

eft

eadem

menfura ac anguli conta&us curva

;

&

tangente con-

tent!

&

fimiliter

curva a circulo ofculatorio infle&itur

per variationem curvaturae fuae, cujus variationis

eft
ri Fig. 12.

eadem
per-

menfura ac anguli conta&us curva

comprehenfi.

& circulo ofculato-

Occurrat recta

TE tangenti DT
Er
3
,

pendiculars curvae in

E &

circulo ofculatorio in r,

&

variatio curvaturae erit ultimo ut

fubtenfa anguli

Contatus

tacts

EDr detur DT) cumque EDr fit Er ultimo ut DT

fi

dato angulo conut ex Art. 3690

traclatus de fluxionibus colligitur, generaliter curvaturae

variatio erit ultimo ut

-_,

Er

Utimur

circulo ad cur;

vaturam aliarum figurarurn definiendam

riationem curvaturae
eft,

verum ad vaaliqua conica,

indefinitis

menfurandam, quae
parabola vel
fectio

in circulo nulla

adhibenda

eft

Quemadmodum autem
qui curvam datam in

ex circulis numero

pun&o

dato contingere poflunt,

unicus dicitur ofculatorius qui curvam adeo intime tangit ut nullus alius circulus inter

hunc

5c

curvam duci

point

bent

omnium parabolarum quae eandem hacurvaturam cum linea propofita ad puncturn datum
;

fimiliter

(funs

Proprietatibus
(funt

generalibus.
infinitas)

385
ea

autem hae quoque numero

tangit

fimul habet curvaturae variationem, quae,

eandem non folum

arcum curvae

& ofculat,

fed

adeo premit ut nullus

alius arcus parabolicus duci poflit inter eas, reliquis

om-

nibus arcubus parabolicis tranfeuntibus vel extra vel intra utrafque.

Qua
ex
iis

vero ratione haec parabola determi-

nari poflit,
intelligitur.

quae alibi fufius explicavimus facile

Sit

DE

arcus curvae,

genti

perpendicularis, fitque rectangulum

DT tangens, TEK recta tanET x TK

DT, &
curva

Temper aequale quadrato tangentis

locus puncti
in S,

SKF

K,

qui rectas

DS

curvae normali occurrat

quemque

tangat in S recta

SV

tangentem
curvaturae

TD fejuncta

cans in V.

Recta

DS

erit

diameter circuli ofculatorii,

;

&

bifecta

DS
9

in f, erit

f centrum

autem

Vf

ft

angulus

SDN

conftituatur aequalis angulo

fVD
git in

ex altera parte refbe DS,

&

recta

DN circulo
&
contin-

ofculatorio occurrat in

turn parabola diametro

parametro

DN defcripta,
erit

quaeque rectam

DT

D,

ipfa erit cujus contactus

cum

linea propofita

in

intimus

atque

maxime

perfectus feu proximus.

Omnes autem
latorii

parabolas alia quavis chorda circuli ofcu-

tanquam diametro

&
in

parametro defcriptae,

&

rectam

DT contingentes in D,
Newtono
in opere

turam cum linea propofita

vaturae a

eandem habent curvapuncto D. Qualitas curenim ut

flu(ft

pofthumo nuper edito expli-

cate eft potius variatio radii curvaturae; eft

xio radii curvaturae applicata ad fluxionem curvae, vel

R
ut

denotet radium circuli ofculatorii

&

S arcum curvae
ut radius

.
S

Ipfa r

autem curvatura

eft inverfe

R,

&

R
variatio curvaturae ut
r*.

RRS

quae eft menfura an^

guli

386
guli

Z>LinearumGeqmetricarum contaclus curva & circulo ofculatorio content!,;

ex
altera" data" facile

Harum autem una

gens anguli

derivatur.

Vatan-^

riatio radii curvaturae in

curva quavis

DE

eft ut

DVS

vel

DVf,

&

in parabola quavis

eft

femper ut tangens anguli contend diametro per pun-

clum contahis
culari.

tranfeunte

&

recta ad

curvam perpendi^

Haec ex theoremate fequenti generali deduci

pofiunt.

Fiff.

4-

torii

I 7'

Theor.

III.

Sit

punclum in

linea quavis

geometrica datum, occurrat

DS

diameter circuli ofcula-

per

ducla curva in

tot punclis

D, A, B, &c.
curvam
contin-

euot

ipfa eji

dimenfionum

ducatur

DT

gent in

D, qua curvam

fecei in punclis I,

&c. binario

paucior'ibus,

&

occurrat tangentibus

AK, BL,

&c.

in

K, L, &c.

eritque variatio curvatures, five

menfura an-

guli contaclus curva

recle ut exceffus quo

& circulo ofculatorio

fumma

comprehenf, di-

reciprocarum fegmentis tan-

gentis

DT punclo contaclus D
&
e.

&f tangentibus

AK, BL

&c.

terminatis fuperat

fummam

reciprocarum fegmentis

eadsm punclo

curva terminatis,
ut

&

inverfe ut radius

curvature

i.

-^ X Jg. + JL + fefc -
Dk curvam
fecans in
<?,

Ducatur enim

recta

/,

&c

circulum ofculatorium in

fitque angulus

iDT quam

minimus
recta

hujus fupplementum ad duos rectos bifecetur

linear

Dab, qu<e

geometries propofitae occurrat

in punctis

D,

a, b,

&c.

&

dufoe tangentes ak, bl,

I,

&c.

fecent reclam

Dk

in punclis k,

&c. eritque per pro-

pofitionem c

praecedentem r

UK

=r

4De
'

=rr
\jt

DA
777

57'

&c

Unde

ik - k

(five

wnbs) =

13?

Proprietatibus general!bus.

387
reclis

'

YT7

&

c*

Pfoinde

coincidentibus

Dk & DK,

feu evanefcente angulo

ffiK,

erit

ultimo
Sit

DR X
~

aequalis ^-= De ~^~, DI

r- DK DL
in
fit

&c.

erT

perpendicularis tangenti
ofculatorio in r
-

T,

atque occurrat circulo

cumque
R<?

re ultimo ad
re

R*

''

alt Ultlm
.

KOTol = 5R-x7F =

eT ad X DS DR X DT>
ut
-re

five
sJ L>

Menfura autem anguli contactus rDe

curva

&

circulo ofculatorio contenti, five variatio cureft ut

vatune,

g-

adeoque ut

xm-m1 1

gj-,

&c.
18. Variatio autem radii curvaturae, live hujus quaa Newtono defcripta, ex priori facillime colligitur.

litas

Junclis

enim

SI,

SK, SL, &c.

erit haec

variatio radii

ofculatorii ut excefTus

DKS,
verfus

quo fumma tangentium angulorum DLS, &c. fuperat fummam tangentium anguCrefcit autem curvatura a punclo

lorum DIS, &c.

e,

&

minuitur radius cfculatorius, quoties arcus

De
ITIc

tangit circulum ofculatorium

DR
&c
*

interne, vel
at contra

cum

4~ "fTT 4"

^c

^"
*

u P erat Tvf

4"

minu-

itur curvatura a

verfus e 9

&

augetur radius circuli

ofculatorii, quoties arcus curvae

De

tangit

larem externe vel

tranfit
fit

intra circulum

&

arcum circutangentem

adeoque

cum

DR

ultimo minor quarn

De

vel

cum

~ + &c.
ut

fuperat

--

+ -I- + &
&c &c.
-

c.

19. Sumatur

igitur in tangente

DT recta DV ita

177

= xk +

m + Cc -

y-y

ui

&c un s - &c. jungatur

->

atur

/v,

388 _Z>LinearumGeometricarum
fV,
conftituatur angulus

SDN

aequalis

DVy^ atque
;

DN N & parabola diametro DN defcripta, cujus parameter DN, qusque reclam DT contingit in D, eandem habebit
occurrat recta
circulo ofculatorio in
eft

variationem curvature
in puncto

cum

linea geometrica propofita

D.

Ex

iifdem principiis alia quoque theorevariatio curvaturae in lineis

mata deducuntur, quibus

geometricis gencraliter definitur.

20. Ut haec theoremata ad formam magis geometricam reducantur, lemmata qusdam funt prsemittenda,
quibus doc~trina de divifione rectarum harmonica
plior

am-

&

generalior reddatur.

In recta quavis DI, fum-

fegmentis DF & FG, ducantur a puncto V quod non eft in recta DI tres recuse VD, VF, VG, & quarta VL ipfi DI parallela, atque has quaptis aequalibus

quovis

tuor rectae, a CI.

tur.

D. De

la

Hire, Harmonicales dicun-

Recla vero quasvis, qua? quatuor harmonicaliOccurrat

in

bus occurrit ab iifdem harmonice fecatur.

recta

DC

harmonicalibus

VD, VF,

VG & VL
ipfi

pundis D, A,
rallela, quae

B C
3

eiitque

DA ad DC ut AB ad BC.
DI
pain

Ducatur enim per punctum

occurrat redis

A recla MAN
erunt
(five

VD h VG

M & N;
ad

&

ob

aequales

Eft autem

adeoque ut

DF h FG, DA ad DC AB ad BC.

aequales
ut'

MA &
AN)

AM
eft,

AN. VC,

Manifeftum

eft

rectam, .quae

uni harmonicalium parallela

dividi in aequalia feg-

menta a
ipfi

tribus reliquis.
reliquis

Occurrat recla

BH

parallela
;

VF

VG, VC,
tit

VD

in

B,

K&H

erit-

que

VK

VK ad

KB FG (vel DF) ad VF KH, & proinde BK = KH.

ad
fequitur,
fi

adeoque ut

21. Hinc

recla quaevis a quatuor rectis

ab eodem puncto duclis iecetur harmonice, aliam quamvis

;-

Proprietatibus- generalibus.
iris

389

re&am

quae his quatuor reHs occurrit harmonice

;

fccari

ab iifdem

earn vero quae parallela eft uni quatuor

reaarum
Sit

in Tegmenta aequalia dividi

a tribus reliquis.

BC, jungantur VA, VB, DA occurrant reds MAN, DFG ipfi VC VC & VD ; parallels reals VD, VA & VB in M, A, N & D,
ad
ut

DC

AB

ad

F,

G;

eritque

BC V
z=z

adeoque ut
;

MA ad VC ut DA ad DC, vel AB ad AN ad VC MA = AN, & DF

;

FG

&, per praecedentem,

reta quaevis quae ipfis

VD, VA, VB* VC

iifdem.

occurrit harmonice fecabitur ab

22.
reclas

Ex

punclo

ducantur duae

rec"tae

DAC, Dae
atque a,
c

Fig. 16.
n.
i,

VA & VC
Ac

fecantes in punclis

A,

junclas

mutuo occurrant VQ_ harmonice fecabit re&am DAC

fibi

& aQ

in Q^,

&

ducla

vel aliam

quam-

vis

re6tam ex punclo

D ad

eafdem rectas ducram. Secet

enim VQ^recliam
recta

AC

in B,
ipfi

&

per

punctum Q^ducatur

DC, quae occurrat re&is Da, VA & VC in punftis M, R & N cumque fit MR ad MQ^ut DA ad DC, & MQjid MN in eadem ratione, erit quoque RQ_ad QN ut DA ad DC. Sed
parallela
;

MQN

RCt eft
ut

ad

QN

ut

AB
la

ad

BC.

Quare

DA
I.

eft

ad

DC

AB ad
23.

BC.

Haec

eft

Prop. 20ma, Lib.

fe&ionum

eonicarum CI.
Sit

De

Hire.

DA

ad
vel

DC

ut

AB

ad

BC,

eritque

-|-

asqualis 1

fummae

differentiae

ipfarum r

& - DA DC
-

prout puncla
punc~ti

A & C funt

ad eafdem vel contrarias partes

D. Sint imprimis puncta A & C ad eafdem partes punai D, cumque fit DA xBC^DCx AB, i. e. DA x

DC DB = DC x DB DA, vel D A x DB DC s= DCxDA DB erit 2D A X DC = DA X DB + DC Cc 2

590

I>LinearumGeometricAruM
X DB,
2

DC
n. 2

adeoque y-g- =z yrr-

11 +

-~p.

Sint

nunc

&

3.

puncta
vel

A & C

ad contrarias partes puncti

D,
vel

eritque

DAxDB^DCrzrDCxDB + DA, DB + DC = DCx DB DA, adeoque

D,

^~
punto

DA X

cum punch B
=.
-pr-r-

vel -p-rJDj5

211
LA
illis

& C funt ad eafdem partes

-^- quoties punfo.

pun&i
funt

jJL

A&B

ad eafdem partes puncti D.

Si igitur, datis

&

rec~tis

VF & VC

pofitione, ducatur

ex pun&o

D
1

recla quaevis

occurrens in puntis

A &
2

C,

5c in
4-

eadem
4?

recla fumatur femper

DB

ita ut

=-=-

&l>

-=-

DA.

j^i U<^

ubi

fupponitur

terminos -rr-

DA &

-=jpr

iifdem

JJL

vel contrariis fignis afficiendos efle prout puncla

A&C
lo-

funt ad eafdem vel contrarias partes puncli

D, erk
;

cus puncti
rectse

ipfa

harmonicalis
fecat in

VG qua? reclam DFG

ita

VC parallelam
per

FG = DF

quseiif-

que

tranfit

punclum Q_ubi (ducla Dae qus

in a et c) junclae

dem reelis VF & VC occurrat t sQ fe mutuo decuffant.

in
-

Ac

Fig. 17.

2 4^-y

Si

recta

DA
in

fumatur femper

Db

ita

ut

J)b

=c t- "Z T

DA

DU

ducatur

DF

parallela r

rectae

VC

quse reels

VF occurrat

F,

qus

reels

VC

occurrat in
;

H,

& DH parallela rectse VF & ducta diagonals HF

1

erit locus puncti b

nam ex
cum HF,

hypothefi
fit

^7

= Tyo
B
fint

&
erit

DB

z=

iDb

adeoque

VG
fi

locus puncti

punctum

b ad rectarn

puncla

A & C

ad

eafdem.

'

Pr-oprietatibus

generalibus.

391
=-7

cafdem

partes puncti

D.

Si

autem fupponatnr

~
vc

yr

jy^9 eadem conftruftio inferviet pro determib,

fi

nando pundto
rects
fed

fubflituatnr loco reclae

VC
a

alia

VC

parallela

ad aequalem diftantiam

puncto

ad contrarias partes.

25.

Ex

punclo dato

ducatur

rec~ta quaevis

DM
A,
4

quae tribus rectis pofitione datis occunat in puncris

C, E

&

fumatur femper

DM
DE

ita

ut

= 4DA
JL>M
-

Trt -r -rr DC r Da,

(ubi termini funt contrariis nVnis afficiendt 6

K

quoties rectae

DA,

DC

vel

punfti

D)

fupponatur

-f,

funt ad contrarias partes

=z

-^p
;

eritque

ad rectam pofitione datam per praecedentem

adeoque,
ad po-

cum

fit

- =

|-

erit

punclum

fitione datam, per

eandem.

Compofitio autem proble-

matis facile ex diHs perflcitur.

tres reclae pofitione datae,

mum DFVH,

ducendo

Sint VA, VC & vE & compleatur parallelogramDF & DH rectis VC k VF

refpective parallelas,

&

cccurrat redla

vE

diagonali in

v ; deinde compleatur parallelogrammum Dfvb ducendo rechs Df & Db reclis vE Sc HF parallelas quae
re&is
nalis
recTtis
1

HF & vE KF &
1

occurrant in pundlis

f Sz b;

&c diago-

bferk locus puncli

hf'm
1

M.
;

L&

DM~I)L + i5E~lJA + DC + Uf
ilrucYio

iii
3

Occurrat enim recta

DA

eritque, ex prsecedentibus,
.,.

C0U ^

ex Art, 22. deducitur.

Cc

26.

'39 2

-D^LinearumGeometricarum

reetis pofitione datis in punctis

ducta occurrat: 26. Recta quaevis ex punclo dato A, B, C, E, &c. et in

hac reaa fumatur femper

4!

^ = -^

+ gg + gq

r,

&c.

eritque locus puncri

M femper ad rec~tam
modum
prece-

IJEj

pofitione

datam.

Demonftratur ad

de ntis.
Fi 18.

2 7* Theor. IV. Circa datum punclumV remlvatur

recla

PD

qua

occurrat linetz geometric^ cujufcunque or-

dims in
fionum^

tot

punSfis

D, E,
refia

I,

&c. quot

ipfa eft dimen-

& ft in eadem

fumatur femper

PM

ita ut

ik i^ + ik +
rum regulam fapius
erit locus punfii

ik +

c'

W
l

figna

termin -

defcriptam

chfervare fupponimus)

line a reIa

Ducatur enim ex polo

recta quaevis pofitione data

PA,
quot

quse curvae occurrat in tot punclis

ipfa eft

A, B, C, &c.
rec~r.ae

dimenfionum.

Ducantur

AK, BL,
et per

CN
Art

curvam

in his punclis contingenfes, quae cccurrant

re&ae

PD

in totidem punclis

K, L, N, &c.
-

Ia

To + T +
&c.

&c Ti +

piT

+ 7L +

- "ZjT

Unde r-

aequalis eft huic

fummse, cumrecuse

que pofitione detur recta PA,

&

maneant

AK

BL, CN, &c. dum

tur, erit

recta

PD

circa

polum P

revolvi-

punctum
;

M ad lineam rectam,
&c. determinari

per articulurn

prsscedentem
oentibus

quae per fuperius often fa ex datis tanpoteft.

AK, BL,

28. Sicut recta

luas rectas

Fm medium

eft

harmonicum
\

inter

PD h

PE, cum -^-

=~

r;

G-

xniliter

Proprktatihts generalibus.
militer

393
n9

Paz dicatur medium harmonicurn inter reftas

quaflibet

PD, PE,
==

PI, occ.

quarum numerus
Et
*

eft

curn
dato

Wn Td + ffi + P recta quaevis du&a

H + &c
&

fl

ex

puna

lineam geometricam fecet in

in

tot punctis quot ipfa eft

dimenfionum,
inter

qua fumatur

femper

Pm medium

harmonicurn

Tegmenta omnia
erit

dutae ad

punctum datum P

curvam terminata,
Erit
;

punctum r

ad rectam lineam.

enim

=--

PM

== -
P//V

adeoque Y?n ad

PM ut n ad unitatem
Atque hoc

cumque punctum
erit

M
vel

fit

ad rectam lineam, per pnecedentem,

eft

quo-

que ad rectam lineam. eidem anine.

#, , f,
*/,

theorema

CrAefii,

29. Sint
ny

&c. radices requationis ordinis

ultimus ejus terminus

ingreditur,

quem

ordinata feu radix y

non

coefficiens termini penultimi,

dium harmonicurn
-]

inter

omnes

radices,

feu

M me = a
JVl

1-

-}-

&c.
c,

Cum
&c.

igitur

fit

factum ex

radicibus

omnibus

a, b,

in fe

que

P fumma

factorum

cum
erit

radices

mutuo ducYis, fit* omnes una dempta

-4'

in fe

mutuo ducuntur, 7

P z=

V
a

V a
a

V
-j
'

V
1-

-7

+
fit

&c

=
"
t

-,

adeoque
cujus

M z= ~.
duse

Sic,
fint

fi

sequatio
,

quadratica,
zz:

radices

et

erit

(affumpta aequatione generali ftctionum

1.

conicarum Art.

propofita) r r '

z= *

2.CXX

2iix

-\~~

2
.

ax

-;

In
erit

b
b,
<:,

sequatione cubica cujus tres radices funt #,

M =

2_ao
-+-

ac

be

(fi

aflumatur

asquatio

ge-

-\-

neralis

394 D^LinearumGeometricarum
neralis
Tfx*

linearum
%

tertii

ordinis

ibidem propofiti)

:=-"

^gx 4- $hx ik cxx dx -^

e

Fig-

9-

3* Occurrant re&ae quaevis duae

punclo

Fm

Sc

ductas, lineae geometricae in punctis

z,

P^ 5 ex D, E, I,
termi-

&c.
inter

et d, e,

Sec.

fitque

Pm medium harmonicum

Tegmenta prioris ad

pun&um P & curvam

nata,

&

P//,

medium harmonicum inter Tegmenta fimilia

;

pofteiioris retae

juncta y.m occurrat abfciffe

-

AP
.

in

H,

eritque

PH

= nVx vel
V
curvam
;

PH

ad

Pm

ut

ad

V *

Secet

enim
tionis

abfciffa

in tot

pun&is B, C, F,

Sec.

quot

ipfa eft

dimenfionum

cumque ultimus terminus

aequa-

(i. e.

V)

fit

ad

BP X CP x FP X
,

&c. in ratione

conftanti, ut fupra (Art. 5.) oftendimus, erit (per Art. 8,

V x _ x V =BP + TF +

x FP

_i_

c &C

.
'

ade q

^ PH

=BP*
(q uo

if + ^ + &c == ^p re&a PM fY = X ~.
N
i

& PH

P;/z

V J
cis eft
tertii

PH

ad

?m

ut

tf^

ordinis ut cxx-

dx

In fedionibus conir

ad 2r
e -f-

d;

Sc in lineis
-{- .

ad

3/a-a-

igx

31.
ftratio

Si defideretur propofitionis praecedentis

demon-

ex principiis pure algebraicis

petita,

ea ope fe-

quentis Lemmatis perfici poterit.

Sit abfciffa

AP
-j-

=
~z

a*,

ordinata

PD

=z

y, ultimus terminus aequationis

lineam

geometricam
-j-

defihientis

V z=. Ax

-\-

Bx n

Cxn ~ *
-f*

&c. penultimi coefficiens

-f-

P == ax v

~~'i

4- bx n

ex

n ~~>

&c.

et

fit

Q_

quantitas

quae formatur du-

cendo terminum quemque quantjtatis

in jndicem ip-

nus

PrGprietatibus
fius
ss

generdibits.
i.

395
e.
fit

in

hoc termino
1

&
1

dividendo per x,

<i= wA^-|-

+
Dp
quae

X Bx n ~ 2
quam

+ ^2 X Qx ~
n

&c, (quae

V
-r-

ipfa eft quantitas

dicimus).

Du-

catur ordinata

angulum quemvis datum ApT)

9

PD pD et P/> ut & tranfmutetur aequatio propofita ad ordinatam u & abfciiTam s ;'&

cum
abfcifla conftituat, fintque recuse
/,

datae

r et

dicatur/>D :=#, A/>z:z,

sequationis
<y

novas,

erit aequalis ipfl

"

aequalis

+
,

Qi

+ P/
fit

cum fit z == AP, terminus ultimus V, penultimi autem coefficiens p erit

.

Cum
y
z=z

enim
;

PD (=^)
P/>

ad j5D

(= a)

ut / ad r, erit
r,

?p =

fit

autem

ad ^>D (z= #) ut i ad

erit

& AP

= = Ap + P^ = z 4-
a_y

His

autem valoribus pro

co-ordinatarum z

et

fubftitutis in aequatione pro-

pofita lineae geometricae, prodibit

et u definiens.

aequatio relationem

Ad

hujus ultimum
fufficit

terminum v

Sc

penultimum pu determinandum,

hos valores fubftituere in ultimo V,

aequationis propofitae, atque

&

penultimo Py,

terminos refultantes colli-

gere in quibus ordinata u vel non reperitur, vel unius

tantum dimenfionis

horum enim fumma

x
ipfius valor
1

dat pu,

il-

Jorum

v.

Subflituatur pro

+
n
x

ku

in

quantitate

vel

Ax n

termini

refultantes
Z Bz,"- hi

+ B**- + (V" + &c. et nAs^'~ ku ~ + Az + Bz -f Bz

2
1

X -Z

+
.

Cz"~ 2 +

Cz 2 X

-iu
r
f-

cc.

foli

ad rem faciunt de qua nunc agitur.

deinde

Subvalor

flituatur

pro x idem valor,

5c

pro y ipfius

396 I>LinearumGeometricarum
valor
-f-

in quantitate

Fy == ax n

"

-j-

bx n

~~ z

+ ^-?i

&c.

X
"

Ji
*~" 2

&

termini
cz
n ~~3

refultantes

foli

az

n ~~ l

^z

&c.
z=i

lu

funt nobis re-

tinendi.
fit

Supponatur nunc %

x9

&

fumma priorum fumma ==

-

squalis

-f-

~~^eft

&

pofteriorum

Uncle manifefium
novae v

ultimum terminum
i

asquationis

V = ,/

'*

&

penultimum

/>#

P/

Qi X u.
inter Teg-

32.

Sit

nunc Pzs medium harmonicum

et

menta PD, PE, PI, &c.

(Am fecet abfciflarn in

P^ medium harmonicum
&c. ut
in Art. 30. juncta
P//.

inter Tegmenta Fd, Fe, Pz,

Hy
s
/x*
.

8c

fupponamus

ordinate

pD

parallelam

effe.

Ducatur
;

^ abTcifiae
ut k ad
,

parallela,

qus
ad
fit

reels

Fm
?i<v

occurrat in

eritque Fs ad Fy, ut

PD
,

pD
_.

vel ut / ad r, et
.

ad

P^

r.
.

Cumque

Fa
ms

= (per Articulum prscedentem) KV v p

.
.

W fiq

ent

Fm aVQi
tp,

_.

-r Fs

nV
i?

- +
.

= nV
m>i
ttt /r ~~ P
*t

rNl
P/

+ ^ """
PH,
i.

Eft autem ms ad sp ut P2 ad

e.

P X P/+ Q
ad

-
FlQk
vel

ut

Fm

ad

PH

adeoque

P
ad

x P/Q^
P
ut

Cc

Fm

ad

PH,

PH
fit

= Fm X ^

vel

~.

Cum
/,

igitur valor reels

PH

non pendeat

a quantitatibus
erit

etr;

fed, his mutatis,

Temper idem,

punelum
aliter

p ad reelam
ftendimus.
^\rt.

pofitione datam, ut in

Theor. 4.

29. alia

Quin & valor recuse PH is eft quem in methodo definivimus j 5c recta Hm omnes

Proprietatibus
pes reclas ex

gener dibits

3 97

P du&as

fecat barmonice,

fecundum de-

ftnitionem feciionis harmonicas in Art. 28. generality

propofitam.

^^^^^^^^2*^2^^^2fe*S^^fe^c*sc^^2
S

EC T

II.

De

Lineis fecundi ordint s^ five feclionibus

conicis.

33.TTpX

iis

quae generaliter de lineis geometricis

1/

in feclione

prima demonftrata funt, fponte

tertii,

fluunt proprietates linearum fecundi,

&

fuperi-

orum ordinum.
bafis
eft

Quae ad

fe6tiones

conicas fpec~tant
quae figura
praeceilluftre-

optime derivantur ex proprietatibus

coni.

circuli,

Verum

ut ufus

theorematum

dentium

clarius pateat,

&

figurarum analogia

tur, operae pretium erit

praemiffis

harum quoque

affectiones

ex

deducere.

Doclxina autem conica de dia-

metris,

earumque

ordinatis (quibus parallelae funt rectae

fectionem contingentes ad vertices diametri)

rallelarum fegmentis quae
runt,
rectis

&

de pa-

quibufcunque occurfluit

&
&

afymptotis, tota
5. oftenfa funt.

facillime

ex

iis

quse

Art. 4.

h fe&ioni conicae infcriptae Fig. 20. 34. Rectae occurrant fibi mutuo in puncto P ; ductae AK, BL,

AB

FG

FM,
per

GN

fectionem contingentes occurrant rectae

in punctis

PE

P dudae

K, L, M,
(

eritque femper

TK Tl =
rat in punctis

m + Tn
D & E)

fl

reh

PE
.

curv3e occur "

= -^r- 7p

Segmentis au-

tem

.398

jD^LinearumGeometricarum
partes pun<5H

tem quae funt ad eafdem

nuntur figna;
Fig. 21. puncti iifque

P eadem
oppofitas

prae pa-

quae

funt

ad

partes
fi

figna praeponuntur contraria.

in P,

Hinc

bife-

cetur

DE

&

ex pun&o

ducatur recta quaevis

fectionem fecans in punclis

&ae

AK

et

BL

Fig. 22. fecent in

K et
erit

A et B, unde ducantur recurvam contingentes quae recTam DE L; erit femper PK -= PL. Quod fi
occurrat, iitque
ipfi

DE

feclioni

non
in

P punclum

ubi di-

ameter quae bifecat reclas

occurrit
-,

hoc quoque
rec*tae
;

DE parallelas cafu PK ~ PL.

et

eidem

fig. 23.

infcriptae in

35. Concurrant punto P

AB

FG

fecTioni conicae

ducantur
et

reclae
fibi

fectionem con-

tingentes in punclis
in

quae

mutuo occurrant

K,

Sc

junta

PK

tranfibit

per occurfum reclarum

quae fectionem contingunt in punctis

reta

et

G.

Si

enim
in

PK

non

tranfeat per oecurfurn

in

reclarum feclio-

nem

tangentium

L; cumque
jdentem, erit

~ +~ ^+^
zz:

et

huic occurrat in
per

illi

prsece-

PL =3 PN

&

coincidunt

pun&a

et

N contra hypothefin.
niutuo occurrere in

36. Eadem ratione patet rectas re&ae tt punclo

tt,

AG et LK
;

BF

fibi

adeoque

pun&a P, K,
datis tribus

efiz

in

eadem

recta linea.
et

Hinc
duabus

pun&is conta&us A,

F,

cum

tangentibus

AK et FK,
&
3

feclio conica facile defcribitur.

Revclvatur enim recta K57P circa tangentium occur-

fam

ut polum, quae occurrat reclis

et
tt
-,

AB

et

FB

in

pun&is P

juntae

Asr,

FP

occurfo fuo
per

G
tria

defcribent fectionem conicam

quae tranfibit

puncta data A,

B F&

continget redtas

AK et FK

in

et

R
37

399 37. Caeteris manentibus, occurrant reclae AF etBG fibi mutuo in punto p, tangentes AK et BL in R,
atque tangentes
et

Proprietatibus. gener dibits.

Fig. 24,

FK

et

GL
m\

in

Q_;
;

&

punta R,

tt,

Q_
n;

p erunt
pundra

in

eadem re&a
et

linea

fimiliter

occurrant

tangentes
Sc

AK

GQJn
77,

tangentes

BR

ct

FK in

P ; m,

p erunt in eadem recta

Art. 35,

linea.

De-

monftratur ad

modum

38. Hinc datis quatuor punclis contactus A, B, F, cum unica tangente AK, occurfus redtarum AB et
et

FG, AF

BG,

atque

AG
P/>,

et

pi et 7r; junclae

autem

Psr,

BF, dabunt pun&a P, et pn fecabunt tanm,

gentem datam
duclae
in

AK

in tribus punctis

et

unde

mG, KF, RB punctis datis G, F

fecYionem conicam contingent

et

B.

et et

39. Datis quatuor tangentibus RK, KQ_, QL, unico puncto contaclus A, occurfus tangentium

LR

RK

LQ_,
et

LR

et

QK dabunt punr,a ?n

et n.

Jungantur

nm ; 5c occurfus retarum LK et RQ_, LK et nm^ RQ_et nm^ dabunt pun&a tt, P et p junclae vero PA, ttA et pA fecabunt tangentes RL, QK et QL in

LK

-,

pun&is contactus B,

G et F.

40. Datis quinque pundlis contaclus A, B, F,

et

G
p

y,
et

junc~tae

GF

et

Gf
et

retae

AB

occurrant in punclis
rectae

P
et

X
-,

jun&ae

AF

x
j

&
et

junctse P/>,

Af occurrant Xx occurfu fuo

BG

in

dabunt puntum

m
in

unde duclse

mA

et

mG

fectionem conicam tangent

&

fimiliter

determinantur rerae quae cur-

vam

contingent in punclis reliquis B,

etf.

41. Dentur quinque

tingentes,

retae fetionem
et

conicam con-

VK, KQ_, QL, hu

uV
}

occurfus tangen-

tium

VK et LQjJabic punctum m-

occurfus tangentium

KQ.

400,

2>Line arum Geometricarum

',

et

KQ_et Lu dabit pun&um n ; jungantur mn 9 LK, VL mu refta LK fecabit re&am mn in P & re&a LV
;

fecabit

ipfam.wa in

X$

juri&a autem

PX

fecabit tan-

gentes

VK et #L in

puncliis contactus

A et B.

Similiter

reliqua puncta contaclus determinantur.

Fig. 25.

42. Datis tribus tangentibus AK, BK^ etRL, cum duobus punctis contactus A et B, facillime determinatur
tertium, per Art. 35.
liquis tangentibus in
fe

Occurrat enim tangens PvL re-

et

L, atque juncbe

AL

et

BR,

mutuo

decuffent in w, juncla Ktt fecabit tangentem

RL
Fig. 26.

in tertio

punclo contactus

F$ &

feclio conica de-

fcribi

poteft ut in Art. 36.

RK cum
p
et
it

43. Dentur quatuor tangentes KQ,, QL, LR et fedfaonis conicae quod non fit unico puncto

in aliqua quatuor tangentium.

Inveniantur

pun&a P,
et 53-D
;

ut in Art. 39.

Jungantur PD, pD,

Sc

pD parallela occurrat recuse RQJn Z j & bifariam fecetur PZ in S & duita p$> fecabit retam PD in E puncto curvse vel occurrat PD recuse RQ_in
ducta

PZ

re&s

%, et

(per Art. 23.) fecetur

PD

harmonice in z

et

E,

Du&a
fint

autem D^r

fecabit

junchm pE

in e, et Ett fes

cabit ipfam

pD

in d, ita ut haec

quoque puncla d9

ad curvam.

Fi".
ll-

27.

44*

Ex

punclro

1.

cHonem conicam
recuse duae

AF

et

K ducantur duae tangentes ad feA et B5 ex pundfco A ducantur AG feclioni occurrentes in F et G

in

juncla

BG
7F ;

fecet

AF

in P, et

jun&a
tt

BF
in

fecet recl:am

AG

in

eruntque

pun&a P, K,

eadem

recta

linea, per Art. 36.

Verum
n. 2,

propofitio haec

generalior

eft.

Si

enim a
fe-

punclo quovis

K ducantur

duse ret?e

KAa, KB

ctionem

Proprietatibas generalihus.
ctionem fec2ntes
in punctis

401
ex panaris
\

A, a

et

B, b

et

et a ducantur reels ad feclionem

autem
riis

BF

fecet

aG
K,
tt

in P,

& ducla

aG juncta G fecet AF in tt ;
et
;

AF

erunt puncta P,

in

eadem

recta linea

quod va-

modis

alias

demonilravimus, unde expeditam me-

thodum olim deduximus feclionem conicam defcribendi Sint A, a, B, b et F per data quaevis quinque puncta.
puncta quinque data
jungantur
;

concurrant recuse
revolvatur recta

ha

et

Bb

in

AF etBF;

PKsr
et

circa

polum

K,

quse occurrat his rectis in

et

P;

duels #P, bn

concurfu fuo

G feclionem defcribent.
Fig. 2S.

45. Sit P punctum datum extra feclionem coniet E cam, unde ducla quaevis fecticni occurrat in

et

fi

ttt

PM

= rrtD
1

4.
'

-r PL

erit

M ad lineam
A et B,
Si
ita

rectam

AB
PA
p

quse feclioni occurrit in punctis

et
fit

utductas

PB, erunt contingentes

in

fectionis.

vero punctual

medio puncto
7
,

reclae

AB

intra feclionem, fitque

pm

=. pd X ^
pe
ipfi

locus puncli ?n erit recta ab per r r

ducla

AB

parallela.

Tangentes ad punch.

et

femper concurrunt in recta

AB,

et

tangentes ad puncta

et e in recta ab,

46. Contingat recta

cantur duae quaevis rectae
currant in

DT feclionem in D, unde duDE

e:

Fig. 29,
*

DA,

n quae feclioni oc-

If

et

A.
j

Occurrat
et duclae
in

DE

reclae

AK

feclionem

contingenti, in
rallelas fecent

K DA

EN,

KM tangenti DT pafumatur in recta

et

M,
Sz

DE,
per
Ct

DR
Art
*

ad

EN
cum
a dt
'

ut

KM
QY*

ad

KE,

circulus ejufdem cur-

vaturae

feclione in

trannbit per R.
*
!

Nam
KE

I5

DV*xDK

DE "" 33K

VExUZ

DR

402 jD^LinearumGeometricarum

DR=^X'^
to.

(quoniam

QV
.

DV

::

2.

XM EN EN DE)' = Quod K ^ fuerit tangens AK parallela retae DE, e. DE ordinata diametri per A tranfeuntis) erit DR = -~r-, ut alibi demonftravivel DR ad DE ut EN* ad DE

KM DK
:

:"

fi

(i.

fi

fit

mus

Art. 373. tractatus de fluxionibus.

Si in

hoc cafa

DE

EN
fit

2
,

diameter,

erit

y-
ut

adeoque

DR,
eft.

aequalis'pa-

rametro diametri
Fig. 30.

DE;

fatis

notum

nem conicam
in E.

47. Ducantur reel's DT, DE, quarum prior fectfoccntingat in D, poflerior eidem occurrat

Ducatur

DA quae
A
;

bifeeet

anguJum

EDT
in

et fe-

ctioni occurrat in

jiingatur

AE,

cui occurrat in
;

reclaDV
ducri

parallela rectse quae

curvam contingit
hasc fecabit
rec~tae
;

VR parallela recuse DA,

fi

A et DE in R ubi

diameter curvaturae

DE eritque DR EDT rectus. Erit enim VR ad AD ut ER ad DE, et ut DR ad DK unde DR ad DK ut DE ad EK, adeoque _L- -J- 1 DR DE
circuius ofculatorius occurrit

angulus

fit

n 2

"ryF*

ut P orte ^ at 5 P er Art. 15.

parallela rehe

Si

autem

fit

tangens

AK

DE

(quo in cafu tangentes

DT
R
E.
et

squales conftituunt angulos

cum

recta

AK et DA quae
punctum
et

proinde perpendicularis

eft axi figurae)

coincident punclra

E,

&

circuius ofculatorius tranfibit per

Sequitur quoque ex diesis refos

EK,

DE

ER

efTe in

progrelEone geometrica.

Fie. si.
et

fecStioni conicas in 4^' Occurrat recta quaevis et E, concurrant rectse curvam contingentes ad

DE

D
E
in

Proprietatibus generalibm.
in

403
ft

punto V.

Sit

DOA diameter
DE EN

per

conftituatur angulus

DVr =: EDO,
Ducatur

erit

chorda circuli ofculatorii.

curvae, & DR (= 2Dr) enim AK fec"tio-

nem
gent!

contingens quae retae

occurrat in

K,

et tan-

EV
;

in

Zj

ducatur
-,

parallela tangenti
fit

DT

DA fecans in N cumque DR ad KA ut EN fitque KZ (= -lAK) ad EK ut VD ad DE, ad EK adeoque' triangula erit VD ad DE ut ^-DR ad EN DVr et EDN fimilia et angulus DVr aequalis angulo
re&am
;

EDO.

Hanc methodum determinandi circulum

non adeo
breviter.

ofcu-

iatorium demonftravimus in tra&atu de fluxionibus,

Art. 375. fed

49. Variatio curvaturae, five tangens anguli contacts fetione conica & circulo ofculatorio comprehenfi, eft direc~te

ut tangens anguli contenti diametro

quae per contaclum ducitur

inverfe ut

&

normali ad curvam,
Sit

&
Fig. 32.

quadratum

radii curvaturae.

enim

DR
erit

diameter curvaturae,
ut

&

haec variatio ad

punclum

ut

-,

per Art. 17.

adeoque,

cum

fit

DV
Quod

ad

Dr
dii

DE

ad

EN,

ut

-^rr-r

EN

Variatio autem ra-

curvaturae eft ut tangens anguli

EDO.

ft

re&a

DO
&

circulo ofculatorio occurrat in #, parabola

diametro

parametro

Dn

defcripta, quseque contingit

rectam
eft

DT in D,

ea erit cujus contalus

cum fecHone

intimus, per Art. 19.

VH circulum
guli

50. Caeteris manentibus, ex puncloV ducatur recta jungatur ofculatorium contingens in ;

Fig. 32,

HD, cumque fit angulus DrV ad reftum, erit

RDH complementum anRDH 3= DVr 2= EDO

;

adeoque variatio

radii curvaturae erit ut tangens anguli

Dd

RDH

404 jD^LrftEARUMGEOMETRICARUM

RDH; &
evanefcit.

coincidentibus reclis

DR

ct

DH

variati

ECT

III.

De

5 l4

Lineis
lineis
tertii

tertii

Qrdinh*

T\^ Jr
\

ordinis five curvis fecundi

generis, uberius nobis

variis

agendum

eft.

Dofere 3

trinam conicam,

modis ufque ad faftidium

traclarunt permulti.
falis

Hanc autem

geometriae univer-

partem,: pauci
effe

adtigerunt; earn tamen nee fterifpero,

3em

nee

injucundam ex fequentibus, ut
praeter

patebit,

cum
non

proprietates
Tint

Ncwtono olim
attentione

traditas, aliae

harum figurarum a plures geometrarum

fupra,

indignae.
tertii

Oftendimus

rectam

fe~

care pofte lineam

ordinis in tribus punclis, quo-

niam
jeales

seqtitionis cubicae tres funt radices, quae

elTe

omnes
tertii

poffiwit.

Recta autem quae lineam

ordinis in duobus punclis fecat, eidem in tertio aliquo

puncto neceftario occurrit,

curvae,

vel parallela eft afymptoto-

quo
:

in cafu dicitur ei occurrere ad diftantiam

infinitam

sequationis

enim cubicae

ft

duae radices fint

recta quae
li-

reales, tertia neceftario realis erit.

Hinc

neam

tertii

ordinis
5

contingit, earn in aliquo punclo-

femper fecat

cum

contactus pro duabus interfectionifit.

bus coincidentibus habendus

Recta autem qua?

curvam

in

punclo flexus contrarii contingit, fimul pro

eft.

fecante habenda

Ubi duo

arcus curvae

fibi

mutuo

occurrent,

punctum

duplex formatur,

&

recta quae alalterura

fecat.

terum arcu-m

ibi contingit in

eodem punclo

Froprietatibus gejieralihus.
fecat.

405
in plu-

Recta autem

alia

quaevis

ex puncto duplice

ducta in uno alio puncto curvam fecat, fed

ribus.

non

52.

Prop.
fecet
;

I.

Sint

dus
tertii

parallels,
ordinis

quarnm
tribus
ita fe-

utraque
punctis
cat ut

lineam

in

recta quae

utramque parallelam

fumma duarum partium

parallels ex

uno

fecantis latere
fit

ad curvam terminatarum squalis

tertis parti ejufdem ex altero fecantis latere ad

fimiiiter fecabit

curvam terminate
his parallelas

omnes

rectas

qus curvs

in tribus punctis oc-

currunt

per Art. 4.

53. Prop. II. Occurrat recta pofitione data lines tertii ordinis in tribus punctis; ducantur

dus

quaevis parallels

quarum utraque curvam

;

fecet in

totidem pundlis

&

folida contenta fub

fegmentis parallelarum ad curvam

fitione

&

rectam po-

datam terminatis erunt

ac folida

eadem ratione fub fegmentis hujus reels ad curvam &

in

parallelas terminatis, per Art. 5.

Has duse proprietates a Newiono olim

crunt.

expofitas fu-

54.

Prop.
tertii

III.

Csteris manentibus ut in

Fig. 33,

propofitione prscedente, occurrat recta pofitione

data lines

ordinis in unico pundlo

A,

&

folidum fub fegmentis

lels

PM, Fm,

P/* unius paral-

contentum

erit

femper ad folidum fub feg-

mentis ^>N, pz, pv 9 alterius parallels ut folidum

AP x

P a contentum fub fegmento Dd 2

AP &

qua*
drato

406 D^LinearumGeometricarum
drato diftantis

bP

puncli

a pundlo

quodam b
ab eodem

ad folidum

Ap x

bp z contentum fub fcgmento

Ap

et

quadrato diftantis puncti

b% per Art.
6.

pun&o
Fig- 34-

55. Prop. IV.

ducatur recta

Ex

dato quovis puncto

linear tertii ordinis

PD qus
eandem

oc-

currat in tribus pundtis

redta

D, E, F,

& alia qusvis

A,
quse
et

PA qus
PD

fecet in tribus pundtis

B, C.
rectas

Ducantur tangentes
inter tres redtas

AK, BL, CM,

occurrant in K, L, et

medium
coredtas

harmonicum
incidet

PK, PL, PM,

tres

cum medio harmonico inter PD, PE, et PF, per Art, 10. & 28.

n. 2.

curvs occurrat in pundtum d ut in Art. 6. & medium harmonicum inter tres redtas PK; PL, PM, erit ad medium harmonicum inter duas redtas PD et
redta

PD

autem unico pundto D, inSi

veniatur

i?d

in ratione 3

ad 2, per Art. 12.

56. Prop. V. Revolvatur recta polum P, fumatur fern per PM in

PD
redta

circa

PD
PD,

squalis medio harmonico inter tres redtas

PE,
per

et

PF,

eritque locus pundti

linea redta,

x\rt. 2,8.

Atque
inventa.

hsec eft proprietas

harum linearum

a Cotejw

Fig. 35.
tii

$ 57- Prop. VI. Sint

ordinis in

tria

pundta lines
-,

ter-

eadem

recla linea

ducantur reels

curvam

in his pundtis contingentes,

qus eandem
fecent

Proprietatibus generalibus.
fecent
in
aliis

407
tria

tribus punctis

atque hsec

puncta erunt etiam in recta

Occurrat recta
punctis F,

linea.

FGH

lineae tertii ordinis in tribus

G,

et

H.

Rectae

FA, GB, HC, curvam

fecent in punctis

in

his punctis contingentes

;

eandem
;

A, B,

C & haec puncta erunt in recta linea. enim AB, & haec tranfibit per C
fi

Jungatur enim
fieri

poteft, ocin

currat curvae

in alio in

puncto

M,

reaae

FGH
7E

P; cumque Ot
per Pr P*

N et ~-pj -f ~ +
tangenti

HC

TX +
quod
Recta

IV

erit
-

pI7'

PN

= PM
et

fieri

nequit

nifi

coincidant puncta

N, M,

C.

igitur

AB

tranfit per
fi

C.
Tint tria

58. Cor oh Hinc

tertii ordinis in

A, B, C,

puncta lineae

eadem

recta linea, duclae autern

AF et

BG curvam contingant in F et G, & juncta FG curvam denuo fecet in H, juncta CH curvam continget in H. Si enim recta curvam contingeret in H quae eandem fenon in C fed ctum cum tribus
caret
igitur fecaret
in alio
aliis

quovis puncto, foret hoc punin

A, B, C,

eadem

recta quae

lineam

tertii

ordinis in quatuor punctis.

Hoc autem

fieri

non

poteft. Incidi

autcm primo inhanc

propofitionem via diverfa fed minus expedita, eandem

deducendo ex Prop.

II.

Similiter

fi

recta

Af
fi

curvam

quoque contingat
juncta

in/*,

& dudta Gf curvae

h.

occurrat in b 9
a punctis
fitis,

Qh

erit

tangens ad punctum

Et

A, B, C,

lineae tertii ordinis in

eadem

recta

du-

cantur tot rectae curvam contingentes quot duel pofTunt, erunt femper tres contactus in eadem recta.

59.
tertii

Prop. VII.

Ex
3

puncto quovis

linese Fig. 36.

ordinis ducantur duse rettee

curvam contingentcs,

Dd

4c8

1>LinArumGeometricarum

tingentes,
fecet

&

re6la contaclus

conjungens denuo

curvam

in alio piihdto, revise

in

curvam
*

in
fe
.

hoc punclo

&
A

primo punclo contingentes

mutuo

fecabunt in punclo aliquo curvs.

ducantur
reclae

Ex

punclo

curvam contingentes
fecet in

in

et

G,

juncla

FG

curvam

eandemque
in

Cpntingat in

H recla

PIC quse curvae occurrat

C,

&

ducla AC

erit

ex

Ccrollario

recla

CA

fit

punclum A. Sequifur praecedente, coincidentibus enim A et B tangens ad pundium A.

curvae tangens ad
1.

60. Cor oh
reclae

Si

ex punclo curvae
in

C
H,

ducantur duae

eandem contingentes

et

&

ex punclo

alterutro

contingentes

AF

et

AG

ad curvam, recla

per con cactus

et

ducla tranfibit per alterum pun-

clum H.

g- 37-

61. eamque
denuo

Carol, 2.
fecet
in

Contingat reels

AC
AF

curvam
et

in

G, duclae autem
et
9

CH

curvam
in

contingant in

recla per contaclus ducla earn

fecet in
fi

Q h
f et

juncla

AG
;

curvam continget

G.

Quod

alia recla

Qh

ex punclo

C
A^

ducatur ad

Curvam

earn eontingens in h

&

junclae
et

b 9

bQ

curtan-.

ves occurrant in

duCtae

Af

erunt

gentes ad puncla/* et g.
F:>. 38.

62. CoroL
duclae

AF

FG
in

fecet

Sit A punclum flexus contrarii unde juncla curvam contingant in F et curvam in H, & ducla AH curvam continget
3.

et

AG

H.

Si

enim tangens ad punclum

curvae in alio

quovis punclo ab
curfu ad

diverfo occurreret, recla ex

hoc ocin

punclum contingeret, quod

ires

flexus contrarii
fieri

A ducla

curvam

A
eft

nequit.

Manifeitum autem

tantum duci

poiTe reelas ex punclo flexus contrarii

curvam

Proprietatibus generalibus.

409
ipfo

curvam contingentes
fimul tangit

praeter earn quae in

hoc

punclo

&

fecat,

atque tres contachis cadere in ean-

dem

rectam lineam.

Ex

folo

puncto nexus contrarii

tres

trcs reclae ductae lint in

curvam

ita

contingunt ut

contatus

eadem rec~H. Sint enim F, G, H, in eadem recta, unde tangentes ductae conveniant in eodem puncto curvae , qucd non fit punctum nexus contrarii ; ducafur ae
e9

curvam contingens
adeoque
rectae

in

a y quaeque

ei

occurrat in

&

juncta ell curvam tanget in H, per hanc propofi;

tionem
in

eH

et

aH

curvam contingerent

eodem puncto H.
63.

e. a, f.

Prop. VIII.

Ex

puncto quovis

linear

tertii

ordinis ducantur

tres redtee

curvam conduos quofvis

tingentes in tribus punclis;

recta

contaclus conjungens occurrat denuo curvas,

&

ex occurfu ad tcrtium contactum ducta curvam

denuo

fecabit

in

puncto ubi recta ad

continget.
ducantur

primum
tres rectas

punhim ducta curvam

Ex

puncto A,

lineae tertii ordinis

Fig.

37

AF, AG, et Af9 curvam contingentes in tribus nunctis F, G, etfj recta Gf quae horum duo quaevis conjungit fecet curvam denuo in N, et recla ex hoc puncto ad tertium contactum F ducta curvam fecet in g turn juncta A^ curvam continget in g. Ducaturenim recla
9

AC

curvam contingens

in
et et

quae

eandem

fecet in

C9

cumque puncta G, N,
tangentes ad puncta

fint in

Cumque

punctum N tranfire per C. eadem recta, tangentes autem FA et NC curvae occurrant in A et C, fitque AC tangens ad punctum A, tangens ad punctum g
(per Prop. VII.) tangentem ad

tranfeant

eadem recta, Sc per A, fequitur

puncta F,

N,

g, fint in

twnfibit per

A.

Dd

64.

410 _2>LinearumGeometricarum
64.
Ccrol.

Kinc

fi

curva defcribatur, ex datis

tres rectae

tri-

bes punclis contaf.us ubi

ex eodem pun&o

curvge duc"be earn contingunt, invenitur

quartum pun-

ctum conta&us ubi recta ex eodem punto curva? ducta Atque hinc colligitur ex eodem curva? earn contingit.
punfco quatuor tantum retas duci poffe lineam tertii
ordinis contingentes praeter rectam quae
in

hoc

ipfo

puncto curvam contingit.

curvas

Si

enim

reclae

ex eodem
punctis

pun&o

duci poffent

earn in quinque
indefinite

contingentes, plures reclse

numero

curvam

contingentes ex

miflis facile colligitur.

facilius

eodem puncto duci poffent; ut ex praeHoc autem Corollarium poftea demonftrabitur. Vide infra, Art. 77.
IX.

Fig. 38.

%. Prop.
tres

Ex pun&o

fiexus contrarii

ducantur

tangentes

ad curvam,

&

recta

contactus conjungens harmonice fecabit reclam

quamvis ex puncto

fiexus contrarii

ductam

& ad

curvam terminatarru
Sit

punetum

fiexus contrarii,

rectae cui-vam contingentes in punclis

AF, AG, et AH, F, G, et H. Ex

puncfo

&

reftam

A ducatur recta quaevis curvam fecans in B et C, FH in P eritque PB ad PC ut BA ad AC,

;

Cum

enim

tres tangentes ad puncta F,

G,

et

H,

in eo-

dem puncto

A conveniant,
adeotl ue

erit

per Prop. IV.

+ 7n
l e
et
-

"Tc^Ta'
eft

m -7c = YP
inter duas redtas

PA
ad
pro-

medium harmonicum

PB

PC

curvam terminatas.

Quae linearum

tertii ordinis

prietas ell fimplicitatis infignis.

66,

Froprietatibui general! bus/

411

66. Corel. 1. Recta quae duas quafvis rectas ex puncto flexus contrarii ductas ad curvam fecat harmonice, fecabitquoque alias quafvis rectas ex eodempuncto
eductas

&

ad curvam terminatas.
2.

67. Carol.

Si recta

afymptoto parallela per pun-

ctum

flexus contrarii dudta occurrat recuse

curvae in

O,

erit

12 r = r

FH

in

R&

adeoque

RA = 2RO.
Fig. 39,

68. Prop. X. Recta duo puncta flexus contrarii

conjungens vel

tranfit

per 3

flexus contrarii vel dirigitur in

punctum eandem plagam

um

cum

crure infinito curvas

Sint

A et a

puncta flexus contrarii, juncta

ha

curvze

occurrat in
Si

a, eritque a

quoque punctum

flexus contrarii.

enim tangens figuraein puncto a

puncto
*,

curvae occuneret in
e,

alio quovis

forent

A, a y

in

eadem

recta.

Verum
Sit

ex hypothefi funt A, a, et a

in

eadem

recta, quae

igitur lineae tertii ordinis occurreret in punctis quatuor.

A punctum
O,

flexus contrarii,

parallela curvae occurrat in

& recta AO afymptoto O ducatur OQ^curvam conAQ_,

tranfibit

tingens in

&

fecans in Q_, juncta

per

ubi curva afymptoton fecat.

6g.
contrarii

Prop. XI.

Ductis

ex puncto

flexus Fig. 38,

tangentibus ad

curvam AF, AG,

AH &
;

duabus fecantibus quibufcunque

et

ABC,
fe-

Afa, junclse Bb
cabunt
in recta

Qc vel Be

et

bC

fe

mutuo

FH quse contactus conjungit.

QA et QC
et BC eidem AB ad AC ut erunt QA, QB, QP et QC,
ipfi
;

Occurrat enim recta Bb

in

FH

in

Q_,
fit

P ; jungantur

cumque

PBadPC,

per Prop. IX.

harmo -

4i2 jD^LinearumGeometricarum
harmonicales, adeoque

jpfam
inde

FH

in

/>,

ita

ut

hb fecabit reclam QC Ab fit ad Ac ut pb ad pc

in c et

&

pro-

erit c

punclum

curvae, per Prop.

et

IX. unde fequitur

converfe redlas

Bb

FH &
;

fimiliter oftenditur reclas

Ce convenire in punclo Qjeclae Be et bC fibi mutuo

occurrere in punclo q ejufdem reclae.

70.

Carol. 1.

Ex
C,
et

punclo quovis Q^reclae

FH

duin

cantur ad curvam reclae

punclis B, b 9

QB, QC,
;

earn fecantes

et

c9

turn junclae

CB, cb,

MN,
et

convenient in punclo fiexus contrarii

jun&ae Be

bC,

Me

et

N, Bb

Cc,

NB

et

MC,

convenient in

reaa FH.
, 71. Corol. 2.

Tangentes ad pundla

niunt in punclo aliquo

quovis

rectas

FH

B et C & a
fi

conve-

punclo

in recla

FH

fito

ducantur tangentes ad cur-

vam, reclae contaclus conjungentes vel tranfibunt per punclum fiexus contrarii, vel convenient in recla FH.
72.
Corol. 3.
c,

Dato punto

flexus contrarii

punclis B, C, b,

ubi duae reclae ex eo duclae

;

fecant, datur recla

FH pofitione

junclae

A, & curvam enim Bb et Cr,

contaclus F,

occurfu fuo dabunt punclum Q_, occurfus dabit q 9 duclaque Q_ ea

& junclarum Be et bQ
eft quae

G, et H, conjungit. His autem quinque punclis datis cum aliis duobus et m determinatur linea tertii or-

dinis quae per haec feptem puncla

tranfit Sc in

A, B, C,

b9

c,

punclo

A &
N,
-

habet flexum contrarium.

M, m Ex
9

punclis

enim

et

dantur
his

N et n
novem
tria
j

ubi duclae

AM et
9

Am

curvam

fecant,
Si

conditionibus linea

determinatur.

autem dentur
w, et s

puncla

M, m
quae
,

et

S$

bec dabunt tria alia

unde darentur undecim

nimiie
funt,

conditiones ad

nguram determinandam,

Proprietatibus
funt.
puncT:is
ctis

generalihts.

413

Similiter dato

punc"to flexus contrarii

A cum
etpunadeoque

F,
et

G, (adeoque

tangentibus

AF et AG)
FG,

quibufcunque, datur recta

et determinatur curva.

puncta

N et ,

in

Contingant reels HB, HC, curvam 73. Cor oh 4. et B et C, et juncra CB tranfibit per A, junhe

Fig. 40.,

CG

FB concurrent in punclo curvam V,

continget in V.
contrarii

et ducta VH curvam Tangens autem ad punclum flexus

recta

PL

A determinatur ducendo AV cui occurrat in L, juncra ipfi AH parallela & bifecanda PL in X

;

enim

AX

erit

tangens ad punclum A.
recise

Occurrat enim

tangens ad

FH

in S

eritque

^ -f -^7 =
1

Jh+Tg-W'
(quoniam

* de l ue

7s+7H=TG--k
in

AC

fecatur
et

harmonice

et

B, adeoque
J-r.
r is.

VA
fi

VF, VP,

VG,

harmonicales)
inter

r
et

E&
unde
in

igitur

PK

medium harmonicum

PS

PH
et

PL

parallela rectae
erit

AH

occurrat rectis

AV

AS

X et L,
74.
ordinis

PX

==

XL.

Prop. XII,

Ex

puncto
rectas

lines?

tertii

jrjg.

4,.

ducantur duas

curvam continin in

gentes in

K,

&

G, juncta FG curvas occurrat tangens ad punctum A fecet curvam

et

jungatur

HM,

cui occurrat

FLK

ipfi
;

AH
turn

parallela in

L,

&
C

fumatur

juncta

HKj

recta quasvis
a rectis
$

FK 3= 2FL AB ex A ducta
erit

haret

monice fecabitur
a curva in B,

HK et HF in
NB
ad

N, F,
ut

ita

ut

NC

BP

ad

PC

Occurrat

414 jD^LinearumGeometricarum
Occurrat enim reta

AB
PA

tangenti
"**

HM in T,
adeo(l ue
i^B

eritque

PB
:=

"

FA

PC

"~~

PT

"^

PC

-p-r-

+ -pTr
Unde

(per

conftru&ionem,

&

harmonice)

=
in

-r-r.

PN

fequitur reclam

punais

etP,

NC fecari harmonice vel NB efle ad NC ut BP ad PC.

fl

75.

Corol. 1.

Hinc

duae quaevis
ita

rec~tae

ex

A duclae

fecentur in

N harmonice
re&ae ex

ut

PC

fit

ad

PB

ut
et

BN, omnes
militer

edu&ae a recHs

HF

CN ad HK

fi-

harmonice fecabuntun
Corol. 2. Si fecet recta

76. eamque
clis.

curva punc"lum duplex non haheat,

in

HK

duobus pun&isy et , duclae

A/et A^
quando

erunt refce curvam contingentes in his pun-

Coincidat enim

punctum

B cum

puncto

N,

pervenit ad

curva; adeoque
-p--,
et coincidit

HK cum f cum ~ + -7^7 = ^t, erit ~? =

occurfum re&as

PN

C cum
Ex

B,

et recta

ex
fi

du&a

cur-

varn tunc continglt.

altera parte,

recta

Af curB
et

vam

contingat tranfibit recta

in

HK

per

f\ ob asquales

enim PB, PC,

hoc cafu, coincidunt puncta

C cum

N.
3.

77. Corol.

Si recta

HK in
AG.

folo

puncto

H curvae

occurrat; duae tantum tangentes duci poterunt a puncto

ad curvam, viz.

AF

et

Quatuor tantum ad
Si

fummum

tangentes duci poffunt a puncto quovis lineae

tertii ordinis

ad curvam ut

AF, AG, A/, A^.

alia quaevis

tangens duci pofiet a puncto

A ad
0, et

enim curvam
puncla

ut Ap, recta

HK

tranfiret per

punctum

quatuor

Proprietatlbus generalibus.
pundta
linese tertii ordinis forent in

415
recta, viz.

eadem

H> /,

g,

9-

E A
-

78. Prop. XIII. Si ex puncto lines tertii ordinis duci poflunt quatuor rectse curvam contingentes, rectas contaclus conjungentes conve-

nient femper in puncto aliquo curvse,

&

recta

quasvis a primo puncto ducta harmonice fecabitur a curva

&

rectis

binos contactus conjun-

gentibus:
Sit A punctum curvae, AF, AG, A/, curvam contingentes in punctis F, G, /,

et

A^,

reclae

et g.

Jun-

gantur
(ex

FG et fg,

quibus occurrat recta quaevis

ABC
;

in P et N & NC harmonice fecabitur in B et P, ita femper NC ad NB ut CP ad PB fequitur ex Corol. 2. precedent Rectae autem FG et fg concurrunt in puncto curvae H; & fimiliter rectae Yf et G^ conve-

A ducta curvamque fecans in B et C)

:

recta

lit

fit

is.

niunt in E, atque
curvae, per

F^

et

G/*in

et

E,

R erunt puncta
eft pofterior

idem corollarium.

Atque haec
tertii

duarum proprietatum linearum

fi

ordinis quas de-

fcripfimus in tractatu de fluxionibus, Art. 402.

recta

Quod

AM
ME,

junctae

curvam contingat in A, et fecet in M, MR, MH, curvam tangent in punctis E,

et

R,

H & re&arum AE
;

HR, AR

et

HE,

AH et RE
et ~

occurfus erunt quoque in curva *.

79.

Corol.

Cum
;

igitur fint rectae

fi

HK, HB, HP,

HC,
linea,

harmonicales
c ;

rectae

HB
Ab

et

HC
c,

curvae occur-

rant in b et

erunt puncta A, b, et

in

eadem

recta

Occurrat enim juncta

* Supple
qtise

curvae in b et c at-

defunt

Scherr.ate,

que

4*6 D^LinearumGeometricarum
que
ipfi

HF

in

/>,

et
effe

HK in n
in

cumque

fit

nc ad nb uf
fi

pc ad pb patet
fit

reda

HC;

&

reciprocc

in

re&a
reri

HC

et b in recta

HB,

erunt

A,

b> c, in ea-

dem
Fig. 42.

80.

Prop. XIV. Habeat

linea tertii ordinis

punctual duplex O.
vse

Ex
rectas
\

puncto quovis cur-

A
-,

ducantur duas

contingentes in
in

AF et AG curvam F et G dudla FG curvam fccQt

OH.
Recta qusevis

jungatur

AB

ex

ducta curvas occurrat in pundlis

B
C,

et

C, recte
ut

FG

in P,

&

redhe

OH in N & re&a NP har;

monice fecabitur in punctis ad PC ut BN ad NC.

Jungatur enim
tangenti

et

ita

PB fit

AO
$

quse recuse
fit

HL

in t

cumque

FG occurrat in p et O punctum duplex, eril

fi+jK=k + k
catur igitur ut

adeoque
t et

/O ad

pA OA,

wH =
O*
ita ut

Afit

Se "

harmonice In
et

pt

ad

p&
et

harmonicales funt H/>, H/,

HO,

HA.
PC

Occurrat recta
"+"

PA

tangenti
5

LH
ent

in

T, cumque

~ +
PB
ad

PB

+ PA ~~ PA + PT = -^ confequenter
j

PC
eft

* PB
NC

-*
ut

PC

ad

BN.
Si

81. CoroL

tangens

HL

occurrat recuse

GZ

ipfi

AH parallels
tranftbit

Z, & fumatur per punctum duplex O,

in

GV= 2GZ, dufta HV

fi

modo

curva tale

pundtum
et

habeat.

Vel

fi

recta

Gra
Ra,

occurrat rectis

AH
m,

HR

in a et r, junctae

rA

et

fe decuffent in

juncla

Hm

tranfibit per

punetum duplex O.

82.

; ;

Proprietatibus generalibus.

417
tertii

82.
ordinis

Prop.

XV. Ex
duse

puncto

lines

ducantur

tangentes,

&

ex alio

quovis ejufdem punclo ducantur redte ad contaclus

tangentes ad

curvam in duobus aliis punctis fecantes hsc duo nova puncta in eodem
curvas convenient.
Fig. 43.

pundo

Ex pun&o A ducantur rebe AF et AG curvam contingentes in F et G. Sumatur punctum quodvis curvae P, jungantur PF et PG curvam fecantes in pun&is K et L; atque tangentes ad puncta K et L concurrent in
pundto aliquo curvae B.
B, ducendo re&am
fecat in

Determinatur autem punclum

quae

PC

curvam contingit

in P, et

fi

enim jungatur

AC occurret
Tint in

denuo curvae

in punto B.

Cum
fequitur

enim puncta F, K, P,

eadem

reta,

&
C

tangentes ad punCta

et

P curvam

fecent in

et

tangentem ad punctum K tranfituram per B. Et ob re&am LGP, tangens ad punclum L tranfibit quoque per B.

83.

Carol. Sint igitur

-

A et B

duo

quaevis puncla in Fig. 44.

linea tertii ordinis

ex utroque ducantur quatuor rehe

curvam

in aliis quatuor puncris contingentes, viz.

5

AF,

AG, A/, Ag et BK, BL, Bi, FL et GK, F/ et G, G/ et F&

B/.
;

Jundse FK

et

GL,

fibi

mutuo occurrent
p
;

in quatuor punclis curvae, P, Q_, q,

&

fi

ducantur

tangentes ad haec quatuor punta, hae occurrent curvae

&

curvam fecat. eadem recla, & ex fingulis ducantur quatuor rectas curvam contingentes in quatuor aliis punctis, recla per duo quaevis
fibi

mutuo

in

pun&o

ubi recta

AB

Unde

fi

fint tria

puncla

lineae tertii ordinis in

punta coatatus

duc"ta

curvam femper

fecabit in alio

aliquo

418

D^LinearumGeoMetricarum
;

aliquo punclo contaclus

Sc

quatuor hujufmodi rectae

per idem punclum contaclus Temper tranfibunt.

Fig. 43.

84. Prop.

XVI.

Sint

ct

puncla duo
reclse

linear tertii ordinis,

ita

fumpta ut

FA

et

GA

curvam

in his punclis contingcntes

conve-

niant in punclo aliquo curvae

curva aliud quod vis

A. Sumatur in punclum P, uncle ducantur

ad puncla
atque

et

G
et

recuse

PF

et

PG
L

quae curvae

occurrunt in

jungantur

FL
iibi

et

GK,

harum

occurfus

Q^

erit

in curva.

gentes autem ad puncla

et

Tanmutuo &:
in

curvae occurrent in punclo aliquo curvae B, at-

que tangentes ad puncla P

punclo curvae C,
in
ita

et

(^convenient

ut

tria

puncla A, B, C,

fine

eadem

recla.

Ducatur enim tangens ad punclum P quae curvae

currat in

00

C,

5c

du&a

AC

fecet

eandem

in
et

& duclae
prsefi

BK, BL,
cedentem.

erunt tangentes ad puncla

L, per

Occurrat recta

LF

curvae in

Qj &

recla

GK
niam

non

tranfeat per

igitur tria

Q_, occurrat curvae in q. Quopuncla L, F, Q., funt in eadem recla,

tangentes vero ad

et

F curvam

fecent in

quitur (per Prop.VII.) tangentum ad

fire

B et A, fepunclum Q_tran-

per

t q 9 in

punclum C. Similiter, cum fint puncla G, K, eadem recla, tangentes autem ad puncla &K

tranfeant per

A et B, tangens ad
Utraque

punclum

q tranfibitquo-

que per punclum C.

igitur recla
q.

CQ^, C^,
Coincidunt
fe-

curvam contingit
igitur

prior in Q_, pofterior in

fi

puncla Q_et q y

enim

diverfa effe

ponamus,
Sint

quitur per Prop. VIII. plures

quam quatuor

tangentes

duci poffe ad curvam ex

eodem punclo C,

enim

A/

Proprietatibus generalibus.

419

Af & Ag

refoe quae

duclae Lf, Lg-,

curvam contingant in f Sc g, Sc curvam fecent in m Sc n\ Sc reclae

Cm, C, erunt tangentes ad pun6ta m et n. Quare haberemus quinque tangentes ex C ad curvam du&as CP, CQ^, Cm, Cn, Sc Qq ; quod repugnat Corol. 3.
Prop.

XIL
Corol. 1.

85;
j?unc"ta

Dato pun&o P, ubicunque fumantur

tangentes ad hasc puncla in curva

F & G, modo
fibi

conveniant, datur pundtum Q_, ubi jun&ae

FL Sc
P

GK
duet

occurrunt

mutuo

Sc curvae.

Et

fi

a punclo

catur reta quaevis

PNM quae
9

curvae occurrat in
;

M,

Sc junc~tae

puncla P, n>

QM, QN, earn fecent in m Sc n erunt Sc m in eadem recta linea. Oftendimus

Sc

enim tangentes ad puncla P

in puncto curvae *.

Q_,

fe

mutuo

deeuiTare

86.

Carol. 2.

Si

fumantur quatuor pun&a F, Gj

Fig- 43

K, L,
cl:a

in linea tertii ordinis, ita ut tangentes ad

Sc

pun-

G,

conveniant

in

aliquo pundto curvae, Sc

tangentes ad punta

K et L,
FL

conveniant quoque in
Sc

aliin.

quo pundto
rent in

curvse, ductae
Sc dudtae

FK
Sc

GL
fibi

concurrent

puncto curvae,

GK

mutuo occur-

pun&o

curvae.

87. Prop. XVII. Sint F et puncta lineas tertii ordinis, ubi li

tur

duo

quasvis

rectee

ducanfecaalia

curvam contingentes,

hse

fe

mutuo

bunt in puncto aliquo curvae.

Sumantur

quatuor puncta curvae L, K, /, g, ita ut duclae LF et conveniant in curva, atque rec~tee Ff

GK

et

G^,

in ea

quoque conveniant

tunc ducfos

L/

et o-K, fe

mutuo fecabunt
* Supple quod

in curva, ut

& duels
Tan*

L-etK/.
deeft in

Schema^e,

420 JD^LinearumGeometricar-um
Tangentes enim ad
in curva, per Prop.
et

pun&a/ et
ut

fe

mutuo

decufFant

XIV.

&

tangentes ad punclra

K
et

L, per eandem.

Adeoque per Corol.

2. prsecedentis ?
et

juntee

/L

et

K^

conveniuni in curva, ut

/K

gL.
Fig. 45.

pofitione

88. Lemma. Dentur tres re&ae IC, IH, et CH, tria puncla F, G, S, quae fint in eadem $

&

recla linea.

Sumatur
jungatur
>

iuncta

QF
P
;

occurrat reclae

HC in
QS

pun&um quodvis Qjn reta IC, IH in L, & juntaQG rec~tse FP, ducta SL occurrat reclis FP et

QP in k et N
fitione datas.

atque puncta k et

N erunt ad re&as porecuse

Jungatur enim IN, quae occurrat

in my ol ducatur per

parallela rectae
in puntis
et

FS
z/,

quae oc->

currat reclis IC, IH, et currat recta

LQ^,

x9

et r

oc-

FG

rectis

IC, IH,

ut

Quoniam N* eft ad Nr ut Ga ad et Nr ad Nu SF ad Sb y erit N# ad N& (adeoque ma ad mb) ut Ga x SF ad GF x S, i. e. in data ratione. Datur igitur

HC, GF,

in

<z,

, et h.

pundum

m, adeoque recta IN/k pofitione

k ad pofitione datam.

& fimi-

liter eft

puntum

si.

2.

quoque puntum
quae retae
rectae

89. Corol. Coincidentibus pun&is S et G, coincidit m cum puncio G. Jungatur igitur IG

HC
in

occurrat in

D,

&

ducla
erit

CF

occurrat
punti

HI

E,

turn juncta

DE

locus

K
Fig. 46.

ubi ductae

GL et FP fe mutuo decuffant.
Sit

9'd ^o-p. XVIII.

PGLFQK
ut

quadri-

laterum infcriptum figure, cujus fex anguli tan-

gant lineam

tertii

ordinis

in Prop.

XVI.

Ducantur re&as curvam contingentes IC,

CH,
Tint in

Hf, eadem re&a

in tribus pundtis
;

Q,

p, L, quse

non

jungatur

IG qus

tangenti

CH

occurrat

Proprietatibus generalibus.
.

42
eadem

occurrat in
rat:

D,

et

HF

quae tangenti CI occur-

in

erunt puncla

D, K, E,

in

recta linea, contingit.

quse quidem curvam in puncto

Supponamus enim polum F, & rectas

rectas

QFL
et

et

FKP

moveri circa

LGP
et

QKG
fi

circa

puncta autem Q_, L,

P, deferri in

polum G, tangenti bus QI,

LI

et

PC

turn

punctum

movebitur in recta

DE,

puncta Q^, L, P, ferantur in curva quae has rectas QI t LI, et PC, in his
per Corol. praecedens.
punctis contingit,
recta

Unde

movebitur quoque in curva

Sed per Prop.

quam

DE

contingit.

XV.

fi

puncta Q_,

L, P, ferantur in linea tertii ordinis propofita, punctum movebitur in eadem, quam igitur recta DE

contingit in

K.

(quae curet 91. CoroL 1. Similiter fi rebe vam contingunt in F et G) occurrant rectas IH (quae

AF

AG

curvam contingit
fecet

in

L)

in punctis

M etNj junctaMP
e^

tangentem

AG

in d,

in e

recla de tranfibit
;

& juncta QN tarvgentem AF per K, & curvam in hoc puncto

D, dy
E, erunt in

continget

atque quatuor puncta

eadem

recta linea.

92. CoroL 2. Ex duobus punctis curvae quibufcunque C et B ducantur ad curvam quatuor contingentes
binae ex fingulis,

CQ_et

CP

ex puncto C,

BL et BK ex
I,

puncto B, fintque harum tangentium occurfus

et

H, E,

turn ductae

cto curvae

F;

LQet EH fe mutuo fecabunt in punatque junctarum LP et ID occurfus erit in

;

puncto curvae

tangentes autem ad puncta

F et

G fe

mutuo
recta

fecabunt in puncto curvae

punctis

quod

eft in

eadem

cum

et B.

93. CoroL 3. Datis tribus punctis lineae tertii ordinis quae Tint in eadem recta, duabus tangentibus ex

&

horum

422 jD^LinearumGeometricarum
horum
Sint
et
fingulis

du&is ad curvam pofitione

curvae

datis, fex

puncta contactus determinantur per hanc propofitionern.

A, B, C,

tria

pundh

data in
et

eadem

reia,

AM

quae prioribus occurrant in

B M, N, e et d; fmtque CD et CE tangentes ex tertio punfc> C dutae atque occurrat CD ipfis BM, BD, AM, et AN, inH, D, h et h CE iifdem in I, E, , et m. His pofitis, junc"ta Ne
BMI,
tangentes ex
9
;

AN tangentes ex A,

BDE,

<-,

fecabit

in pun&o contactus Q_, M^ fecabit CD in punlo contactus P, ID fecabit tangentem AN in pun&o contactus G, EH tangentem AM in contactu F, mh fecabit tangentem BH in L, & de-

tangentem CI

fangentem

r,ique nc

tangentem
Diverfse

BE

in

in hoc cafu determinatum

admittit.

K. Quamvis autem problems fit, folutiones tamen plures

lineae tertii ordinis, fed

enim

mero

definitae,

per tria puncta

A, B,

et

nuC, duci pofTunt

contingentes fex rectas pofitione datas

AM, AN, BM,

tangenti

BD, CD,
/>,

et

CE.

Occurrat enim

Ne

CD
mh

in
9

recta

Md

tangenti

CE

in ^,

ID

tangenti
in
/,

EH
P

tangenti

AN
;

in

gy

nc tangenti
tertii

BM

AM in f
et

tan-

genti

BD
et

in k

atque Iinea
fatisfacit

ordinis quae conditio-

nibus propofitis
in

continget rectas

CD et CE vel

Q_,

vel in

et q.

Ea

continget rectas
;

AM et
autem

AN vel in punctis F et G vel infei g BM et BD vel in L et K, vel in et

/ k.

rectas

Conftat igitur

plures lineas tertii ordinis problematis conditionibus fatisfacere pofle, fed

numero determinatas, adeoque pro-

blema

efTe

determinatum *.
4.

94.
ordinis

Carol.

Datis

duobus punctis

lineae

tertii

et

B, tangentibus quoque

AM, AN,
curvam

BM

BD
et

pofitione datis

cum

tribus punctis contactus F,

G,

L, datur punctum

K ubi

recta

BD

contingit.

* Supple cued defunt in Schemate.

Proprietatibus general!bus.
Si

423

enim ducantur rectae' N? et LF, harum occurfu dafecabit contingentem bitur punctum Q_, & ducta BD in puncto ccntaclus K. Datur quoque punctum P

QG

occurfus rectarum
tres

LG et Md,
M*/,

vel rectarum

M^/etFK $

enim

redrae

LG,
Sit

etFK,

neceflario conveniunt

MedN quadrilaterum quodvis, fumatur punctum quodvis Qjn diagonali Ne et P in diagonali Md, recta quaevis QFL ex Q_ducta fecet latera Me et MN in F et L, duct a PL fecet latus Nd in G,
in puncto P.

jungatur

QG

quae latus de fecet in

in

K;

atque puncta

F, K, P, erunt femper
rius oftenfa.
poflibile,

eadem

recta linea, per fupe-

Unde

conftat problema
tres rectas

non ideo

fieri

imin

quod oporteat

LG, Mi,

et

FK,

eodem puncto convenire.

95.

Prop. XIX.
in

Sint

D, E,

F, puncta

Fig. 47.

lineas tertii ordinis in

eadem

recta, fintque tres

rectae

curvam
ut
\

his punctis

contingentes

fibi

mutuo

parallels.

In recta

P
et

ita

2PF
dz
fi

fit

medium
e9

DF fumatur punctum harmonicum inter PD

per

PE

alia qusevis recta

erit

ducta curvas

occurrat in /, d, et

femper iPf medium

harmonicum inter Yd et P^. Supponimus autem puncta d et efle ad eafdem partes puncti P, punctum autem / efle ad contrarias.
Occurrant enim tangentes
in punctis

DK, EL, FM,

rectae

df

K, L,
-^r-

et

eritque per Art. 9- p-?

nr
ut

t == PM r PK
F<?

PL

f fi
^

recta

Qa

tangenti6
ita

bus parallela harmonice fecet rectam

fit

PD

PE
fd
in

ad

EQ ut

PD

ad

DQ_,

et

Q^

occurrat rectae

Ee

424 X^LinearumGeometricarum

= ^ p- (quoniam P? ad PM ut PQ^ad PF, & ex hypothefi 2PF = PQ^, adeoque 2PM z=L?q) ^Hl-pk^ 05 unde^ = ^ + ^, adeoque
in q)
eft

2?f

eft

medium harmonicum
C<?n?A 1.

inter

P^

et P*.

in puncto

Jungantur D^/et E* quae conveniant V, junctae VQ_ et Ff erunt parallelae 3 Sc prcducta VQ_donec occurrat rt0X.2s.fd in r, erit P/*==

96.

J-Pr.

Recta enim PD fecatur harmonice in E et Q_, ex hypothefi, adeoque etiam recta Yd fecatur harmonice in e et r 9 per Art. 21. unde P/"z= 4P r 5 cumque fit
fequitur rectam

PF =2 |PQj
monicali

parallelam efle har-

VQr.
2.

97. CV0/.
ctual p
ita

Similiter
fit

fi

fumatur in recta

DF

puninter

ut 2/>D

aequalis

medio harmonico

$Yj et />F,

ex p ducta curvae occurrat in fegmentum hujus rectae ex una parte pundti p ad curvam terminatum aequale dimidio medii barmonici inter duo fegmenta eodem puncto p et curva
tribus punctis, erit

& recta quaevis

ad alteras partes terminata.

F'ig.

48.

98. Lemma.

Ex

cen ro gravitatis triangull ducatur

recta quaevis quae trihus lateribus trianguli occurrat,

fegmentum hujus
latere

rectae centre gravitatis

erit

& & uno trianguli

inter
aliis

terminatum

dimidium medii harmonici

fegmenta ejufdem

rectae centro gravitatis

Sit

&

duobus

trianguli lateribus terminata.

P centrum
per
et

gravitatis

trianguli

VTZ

occurrat recta

FDE

ducta latead eafdem

ribus in F,

D, Ej
P;

fmtq'ue puncta
eritque

partes puncti

~ z= +
MPL
lateri

E
.

Ducatur
parallela,

enim per punctum P 2

recta

VZ

quae

Proprietatibus generalibus*
quaa lateribus

42 5

VT, ZT,
ob

occurrat in
;

et
fit

VN
et

parallels lateri

ZT in N

cumque

MP PL,
erit

et recta?

LM
inde
fi

TL 3= 2VL,

fimilia triangula

TLM, VLN,

LN - LP, et PN =: 2P1M, proPD occurrat recras VN in K, erit (per Art. 21.

2LN, unde
p D i" pe

&

2 3J

rK

PF*
tres reclse

99.

Prop.

XX.

Contingant
ordinis

VT,
cen-

Fig. 49.

VZ, TZ, dcm recta

trum
c

lineam

tertii

tranfeatque ea-

linca per tres contactus

& per
recta

gravitatis trianguli

VTZ
a

-,

quasvis

per hoc centrum ducta curvs occurrat in punclo

ex una parte

&

in punclis

et b

ex altera

ejuf-

dem

centri gravitatis parte, eritque

inter

zPc medium

harmonicum

Tegmenta Pa
P<;

et Pb.

Occurrat enim recta

lateribus trianguli

VTZ
/-;

in

f,d,ete;k

rectae

VN lateri TZ parallelae in

erit-

que 2 P/= P*. adeoque

= = 1. + 2. =1^
^zzjl + J-,
inter rectas

+ T6--k + Ff>
Pc
eft

adeoque

unde
et Pb

dimidium medii harmonici

Pa

ioo.

Prop. XXI.

Sit

V
et

in linea tertii ordinis,

in

VT

punctum duplex VZ reclas curvam

in

Fig. 50.

hoc puncto contingentes, quibus

T
F

et
ita

occurrat recta

TZ curvam
FV,

contingens in
in

FT = FZ;
-FV
-,

jungatur
recta

&

fi

qua fumatur FP qusevis per F ductae curvs oc<:,

ut

currat in tribus punclis a, b>

fint

quorum a

et

ad eafdem partes puncli P,

ad partes contr arias,

426 jD^LinearumGeometricarum
trarias, erit

femper 2Fc medium harmonicum

et

inter

Tegmenta Fa ^
enim

P, feu qire

= k- + rb a
if

ttt-

Cum
anguli

bifecetur

TZ

in F, fitque

FP
in

==

|FV,

manifeftum

eft
;

VTZ

punclum P efie centrum cumque fit pundtum P

gravitatis tri-

re&a

FV

quae per contacts tranfit, fequitur propofitio ex prascedente.

erit

01. CoroL

1.

Si

jungatur

rec~fee

Va, Vb,
curvam
du&ae

et

Fr,

P quoque centrum

gravitatis trianguli hifce rectis

in a, b9

contend, ut
cy

&

trianguli tribus reclis

;

contingentibus comprehenfi

&

fi

Va

et

Vb

occurrant feclae Fc in
Fff.

et , erit

femper

Fm

aequalis

parallela retae

102. CoroL 2. Reta. per pun&um duplex ducla Fc harmonice fecabit ipfam Ya in k ita

ut

Fa

erit

ad ak ut Fb ad Pi

quae vero ducitur a pun-

k> k ad x occurfum reclarum

curvam

in a et b continc*

gentium

parallela eft
3.

reh

cy

figuram contingenti in
c

103. CoroL
quaevis ex

Datis duobus puneHs a et

ubi recla

P ducta curvae occurrit, datur tertium b ; jungantur enim Va et Fc quae fibi mutuo occurrant in m ; fumatur Fn ex altera parte punH F aequalis ipfi Fzw, et

juncta

Vn

fecabit

reclam Fa in

b*

g. 51.

I0 4-

quodvis

Prop. XXII, Dueatur per punclum re<5U quse dirigatur in plagam crurum

infinitorum

&

occurrat curvas in punclis a et

dueatur per idem punclum recla qusvis curvam

fecans in punclis

D, F, F, qusque

reftis

curvam
9

in a et c contingentibus cccurrat in k et

atque

afymptotq

Proprietatibus generalibus.

427

afymptoto cruris
k,

infiniti in / 5

&

fi

puncta D, E,

w,

/,

fmt ad eafdem partes pun&i P, pundtum

vero

ad contranas,

erit

-^

p^

._ J_ _-,
eft

ubi termini cujufvis lignum

mutandum

quoties fegment-Jim ad,oppofitas

partes puncti

protenditur.
I.

Sequitur exTheor,

rema

11 +

Art. 9. eft
1
T

enim per hoc theo-

__

jj

-j-

p^

-^
&

-f-

pE

1
..

pjh

105.

Can?/. 1.

Si recta

PD

ducatur per concurfum

t.angentium

fcarmonico inter rectas

erit
-J77

P/

132 =
7^-7

ei

cm;

fumatui

PM

squalis medio
Art. 28.

PD, PE, PF, fecundum

adeoque f s *
\?k.

PM

-rrr,

rk

PM erit medium
fi

har-

monicum
per

inter

P/

et

Quod

tangentes ak et

cm concurrant

in ipfo puncto

M,

afymptotos quoque

tranfibit.

106. Corol
funt in

2.

In cafuProp. XIX. ubi

linea

tres

contaclus Fig. 47.

parallels,
fitque

eadem recta

&

tres tangentes

fumatur punclum

ut in Propofitione

XIX.

a?c

afymptoto

parallela,

occurrant ak et cm tangentes reels

PD

in k et

eritque

-f-

-,

five

P/ soualis
fi

dimidio medii harmonici inter Pi et Pm.

gentes ak et cm concurrant in
erit

Quod

tan-

eodem puncto reels PD,

in Prop.

P/

= ipi
erit

quoniam vero
?r

XIX.

,+

Yp

?a

=?

108.

428 jD<?LinearumGeometricarum
Fig. 49.

107. Corol.

3.

Idem dicendum

eft

de cafu Prop.

XX.

ubi tres contactus

tranfit per

D, E, F,
autem

ftint in

eadem reda quae

P centrum
Si

gravitatis trianguli
altera

VTZ tangenfat'

tibus contend.
vel
c

reciarum curvam in a

contingentium (pofita aPc afymptoto parallela)

retae

DP parallela,

abibit afymptotos in infinitum, erit-

que crus parabolicum.

108. Corol. 4. lifdem pofitis ac in Prop. XXI. Sit cPa afymptoto parallela* occurrant tangentes ak, cm 9

reds

VF

in k et m* eritque -r

= - 4- k
curvae

Unde

ft

curva diametrum habet,

pig. 52.

cum
per

hsec neceflario tranfeat

per

punctum duplex V,

&

pun&um

F ubi bi-

fecatur tangens

TFZ,fumatur abF vcrfus V, FPz= 3FV,

ducatur c?a afymptoto parallela,

&

tangens ak quae dia-

metro occurrat
tur, fuper

in k 9

&

ex

altera parte, pundli

P fuma/

redam PV, P/
fit

= |Pi,
erit

&

re&a per
erit

duda
Si

ordinatim applicatis parallela

afymptotos curva?.

vero tangens ak

diametro parallels,
Propofitio

crus curvae

generis parabolic!.

Newtoni de fegmentis

reels cujufvis tribus afymptotis

fequitur ex Art, 4. ut ab
aliis

& curva terminatis facile

dim
oftenfurn
eft.

F %-

53-

l0 9-

Prop.-

XXIII. Ex punclo quovis

ducantur duas quasvis

lineae tertii ordinis

rec~te

DEI, DAB, quse curvas occurrant in pun&is E, I A, B; ducantur tangentes AK, BL, quas recipe DE occurrant in K et L. Sit DG medium haret

monicum. inter fegmenta

minata, atqiie

DE,

DI, ad curvam

ter-

DH
DY

medium harmonicum
medium geometricum

inter

fegmenta
abfeifla.

DK, DL,
Sit

ejufdem reels tangentibus

inter

PG

Proprietatibus

generalibus.
parallela tangenti
fi

429

DG
dem

et

DH,

ducatur

VQ^

DT,

quas occurrat rectseDA

m Qj &

circuius ejuf-

curvature

cum

linea tcrtii ordinis

propo-

fita "in

punch)
et

occurrat recuse

DE
eft

in

R, erunt

HG, QV
Nam

2DR

continue proportionales.

DE

111 +
(

per Theor. II. (Art. 15.)

1

py^ '^ =
_ 2DH-2DG
~"

I~"*"DK^DL"~I>g"~DH
q uoniamDV
2

CG x DH

~^

=DGxDHO
ad

unde
ad

QV

5tHG X DR, adeoque

10. Ccrol.
1.

HG

QV

ut

QV

2DR.

Sumatur

igitur

proportionaiis re&is
recuse

HG
in

et

Dr in ^QV, &

rec~ta

DE tertia

perpendicubris

DE

ad

punum

r fecabit

normalem tangenti

DT

ad punctum

centro circuli ofcuiatorii five

circuli

ejufdem curvaturae
Si

&o O.
i.

punda E,
prout

ejufdem puncli prout

cum linea propofita, in punK, L, fint ad eafdem partes DH major eft vel minor quam
I,

DG, e. DK, DL
minata.

medium harmonicum

inter

Tegmenta

tangentibus abfcilTa majus eft vel minus

me-

dio harmonico inter Tegmenta

DE, DI,

ad curvam ter-

in.

Corel. 2. Si
et

angulus

EDT bifecetur rec"ta DA,

ad

eritQVzrDV,
adeoque

2HG x DR = DV 2 = DG x DH,
2DR.
rec~ta

HG ad DG ut DH
Corol. 3.

1 12. manente

Revolvatur
et

DA circa polum D,

recta

DE,
et

monicorum

DH

HG, differentia mediorum harDG, augebitur vel minuetur in

duplicata ratione refoe

circuli ofcuiatorii
Jis eft

VQ

Quippe ob datam chordam

DR., manet quantitas

0\/2
quae aequa-

2DR,

430 I>LinearumGeometricarum
Fig. 54.

113. Cor oh 4.
ut

Si

tangentium
parallela,

AK

et

BL

altera,

BL,

fit

reche

DE
X,

ducantur

GX

et

KZ
ipfi

parallelae

recbe

DT

curvam

in

D contingently
2DK

quae

AB
_L

occurrant in

entque

DK ~ DG x~DK' DG DE r= 2DK DG, & proinde erit ut oque 2DK DG ad KZ ita GX ad DR. Si tangens AK evadat quoque parallela re&ae DE (quod in his figuris
1

"T"

_L DI

J_
DK

_L_

_L

m^ = PG
ut

contingere poteft)

erit

DG
x

ad

GX

GX

ad

2DR GX*

nam

in

hoc cafu

DG x 2DR.
D,
tur

Du 1

^-rr

DR.

DG

adeoque *

114. Cor oh 5. Si recla DE fit afymptoto parallela, adeoque curvae occurrat in uno pun6to E praeter ipfum
fitque fimul tangens

BL
ut

afymptoto parallela, ducaquae occurrat reclae

EY
Y,

parallela tangenti

DT

DA

in

eritque

KE

ad

KZ

EY

ad

DR *.

115. Coroh 6. Si fit D puntum flexus contrarii, punclum H cum G, evanefcente linea HG, adeoque evaditDR infinite magna, i. e. curvatura minor eft ad punctum flexus contrarii quam in circulo quantumvis magno ut alibi quoque oftendimus, tra&atus de
coincidet
;

fluxionibus, Art. 378.

F*g- 55

IX ^- Coroh
ptoto parallela,

7.

Sit

punclum duplex,

DA

afym-

&

occurrant reclae
in

VQ_, KZ,

tangenti

DT parallelae recuse DA
afymptoto in L, fitque

Q_et Z, atque occurrat

DV

DK et DL,

DH medium harmonicum inter eritque .2DH DG ad KZ ut DL ad DK,

* Sapple figunmi

atque

Proprietatibus generalibus.
atque

431

VH

HN

VQ_ DR.
:

Si
:

reda
:

DA bifecet
Fig. 56.

angulum

TDV,

erit

DR DV
:

DH

2VH.

117.

Prop.
I,

XXIV.

Sit

punctum quodad

vis lineas tertii ordinis, occurrat tangens

curvse in
torii,

fitque

DS

diameter circuii ofcula;

quae curvae occurrat in

ductse
fit

DH

A et B unde rectacurvam contingenres fecent DI in K et L medium harmonicum inter DK et DL,

&

fumatur

DV

ad

DI

ut

DH

ad differentiam

rectarum 2DI et
turae inverfe ut

DH;

eritque variatio curva-

reclangulum

SD x DV;

& jundla

VS,

variatio radii curvaturae ut tangens anguli

DVS.

Nam
1

perTheor.

III. (Art. 17.) variatio

1

curvature
2

eft

u
1

ds
X

~i

~~

2DI

DS

DH DH xDI

dk

+ dl"~ m ~
1

d~s

x bid
.
.

T
131

__

=
DS

DSxM'

Wnltl 2Utem

dii ofculatorii eft ut

rr,

adeoque ut tangens anguli

DVS,

per Art. 18. parabola autem quae

eandem habebit

curvaturam

&

eandem variationem

curvaturae

cum linea

propofita, determinatur ut in Art. 19,

118. Corol
rallela,
erit

Si

tangens

BL

fit

tangenti ad

pa- Fig. 57.

DV

ad

DI

ut

DK

ad

IK;

&

fi

utraque
erit

rangentium

AK, BL,
hoc cafu
lit

fiat parallela ipfi

DI, adeoque variatio curvaturae inverfe ut DS xDI.

Quod
fi

DT

DV

in

DT parallela

afymptoto curvas, Fig. 58,

igitur
fectio-

evanefcet variatio curvaturae.

Quemadmodum
axium
in

evanefcit variatio curvaturae in verticibus

num

conicarum

ea fimiliter evanefcit

verticibus

diame-

432 D<?Line arum Geometricarum 3 Q?^

diametrorum linearum
tertii

ordinis quae ad rectos an-

gulos ordinatim. applicatas bifecant.

Fig. 59.

SchoL Sunt autem alia plurima theoremata de tangentibus

curvatura linearum

tertii ordinis.

Sint, ex.gr*

et

G duo puncta

lineae tertii ordinis

unde tangentes"

dutae concurrunt in curva in A. Producatur

curvs occurrat in H.

FG donee TAG tangens ad pun&um A & conftituatur angulus FAN = GAT ad contrarias partes retai urn FA, GA, fecetque AN rec~tam FG in N.
Sit
9

Et

fi

circuli ofculatorii

occurrunt

rec~tae

FG

in

et b t

re&angulum NFH ad NGH. Sit enim pun&a a ipfi A quamproximum, & punta/*, g 9 k 9 ipfisF, G,H, quamproxima, eritque AY a FG/: GF
erit

GB

ad

Yb

ut

FB.
:

FG/(= HG)
: :

HFZ>

FH GH.
:
:

HYZ>
: :

(=

GFg) AGa ::bG:GY; unde AF*

AGa
:

FH X

bG FB xGH
:

GN FN
:

unde

FB Gb

NFH

NGH.

Sed de

his fails.

FINIS,

%* *s6

BOOKS printed for A.

rine-Street in

Millar

oppo/iteKtths--

the Strand.
In two books.

An Account of Sir Isaac Newton's Philosophical Discoveries. In four books. Publifhed from the Author's manufcript papers, by Patrick Murdoch, M. A. and The fecond edition. 8vo. F. R. S.
Geometrica Organica
varum
univerfalis.
:

Treatise of Fluxions.
lumes, 4to.

Two

vo-

five defcriptio

linearum cur-

4to.

The above
&c.

three books

by Colin Maclaurin, M. A.

earth.

Mercator's Sailing, applied to the true figure of the With an Introduction, concerning the difcovery
that figure.

and determination of

Newtoni Genesis Curvarum per Umbras.

fpe&ivae Univerfalis Elementa
;

Seu Per-

exemplis coni fe&ionum et

linearum

tertii

ordinis illuftrata.

Thefe two books by Patrick

Murdoch, M. A.
:

&c.

The Doctrine

of

Chances

or,

culating the Probabilities of Events in edition, fuller, clearer, and more correct than the

Method of calPlay. The fecond

firft.

Miscellanea Analytica de Seriebus et Quadra-

ture.
at

Cum

Supplemento.

4to.
feveral Tables, exhibiting

Annuities on Lives: with

one uiew, the values of
tereft.

lives, for different rates

of in-

fourth edition, in which are added, Tables for three and three and a half per cent.

The
laft

The

three

by

Abraham De Moivre,

F.

R.

S.

of

London and Berlin.

tranflation

Euclid's Elements of Geometry, from the Latin of Commandine. By Dr. John Keil and

Samuel Cunn.

The

feventh edition.

compleat Treatise of Practical Navigation,

:

demonftrated from its firft principles together with all the neceftary Tables. To which are added, the ufeful theorems of menfuration, furveying, and gauging; with their application to practice.

By Archibald Patoun,

F. R. S.

The

fourth edition, with additions.

BOO KS

printed for j.

NoufcSE at

the

Lamfc

oppofite

Katherinc- Street in the Strand.

chiefly in-

The Elements of Natural Philosphy, of

tended for the ufe
fludents in

univerfities.

By

Peter Van Muschenbroek, M. D.

lated

profefTor of

matics and philofophy in the univerfity of Leyden. from the laft edition of the Latin by John Colson,. M. A. F. R S. Lucafian profefTor of mathematics and philofophy in the univerfity of Cambridge. Illuftrated with copper plates. Two volumes, 8vo.

matheTranf-

The Elements of Astronomy,

vations. and demonflrated

deduced from obferupon the mathematical principles

of the Newconian philofophy. With practical rules, wherethe principal phaenomena are determined. To which is annexed, A Treatife on Projection in general ; defigned for ftudents in univerfities. The whole tranflated froin the French ofM. de laCaille, member of the royal academy of fciences at Paris. By John Robertson, F. R. S. With additions and corrections communicated by the author. Price 6 s. bound.

by

Memoirs of the Royal Society;

ment of
this

or,

the Philofophical Tranfaclions.

is

A new AbridgIn the courfe of

from the
origi-

work, every thing

carefully extracted
;

nals according to the order of time

the Latin tracts englifh-

to praSice,

ed, the terms of art explained, the theoretical parts applied and the whole illustrated with great variety of

copperplates. volumes, 8vo.

By Mr. Baddam. The

fecond edition. In ten

Mathematical Lucubrations. Containing new improvements in various branches of the mathematics. By John Landen. Price 6 s. few'd.

A Course of Lectures, containing near 200 experiments, fully illuflrating the feveral parts of natural philoBy the late Richard Helskam, M. D. profeflbr of fophy. phyfic and natural philofophy in the univerfity of Dublin, To which are added, by way of Appendix, feveral problems,
by
the editor,

Bryan Robinson, M. D.
Price 5
s.

The

third edi-

tion, in octavo.

bound.

The Mathematical Repository.

lytical folutions

of

five

Containing anahundred queftions, moftly felected

from fcarce and valuable authors ; defigned to conduct beBy ginners to the moll difficult properties of numbers. James Dodson, F. R. S. mailer of the royal mathematical fchool at Cbrijl-Hcjpital. Three volumes, 12 mo.