Quantity and Accent in The Pronunciation of Latin - F. W. Westaway (1913) PDF

<^%CA,^^
CORNELL UNIVERSITY LIBRARY
3 1924 064 122 660
OLIN llBRA^Y-CIRCULATIO^.'
DATE DUE
35!SSfti
The
tine
original of
tliis
book
is in
Cornell University Library.
There are no known copyright
restrictions in
the United States on the use of the
text.
http://www.archive.org/details/cu31924064122660
Production Note
Cornell University Library produced this volume to replace the
irreparably deteriorated original.
It
was scanned using Xerox software and equipment at 600 dots
per inch
resolution and
compressed prior to storage using
CCITT Group 4 compression.
The
digital data were used to create
Cornell's replacement volume on
paper that meets the ANSI StandThe production
ard Z39. 48-1984.
of this volume was supported in
part by the Commission on Preservation and Access and the Xerox
Digital file copyCorporation.
University
by
Cornell
right
Library 1991.
dnrnpU
llniuprBtty SItbrarg
BOUGHT WITH THE INCOME
OF
THE
SAGE ENDOWMENT FUND

THE GIFT OF
1891
^3.^3 ie.3.,.^-
x,a.:>c:..lk
QUANTITY AND ACCENT

IN
THE
PRONUNCIATION OF LATIN
CAMBRIDGE UNIVERSITY PRESS

FETTER LANE,
ILonlion:
C.
ffiUinSurgt):
Berlin:
Itipjis:
^ttai
ombE
anil
B.C.
CLAY, Manaqer
F.
100,
PEINCES STREET
ASHER AND CO.

F. A. BEOCKHAUS
A.
gotk: G. P.
Caltutta:
PUTNAM'S SONS
M.\CMIIJ.AN
All rights reserved
AND
CO., Ltii.
QUANTITY AND ACCENT

IN
THE
PRONUNCIATION OF LATIN
BY
F.
W.
WESTAWAY
Cambridge
at
the University Press
E.V,
..>^'
"Si nimius videor, seraque coronide longus

Ease
liber, legito
pauca; libellus
Terque quaterque mihi

Pagina: fac
tibi
finitur
me quam
ero.
carmine parvo
cupis ease brevem."
PREFACE
riiHE
-*-
correspondence in The Times during the past year
makes
clear that
it
we have
still
of that older race of classical scholars
to
be
Yet
let alone in their
easy-going pronunciation of Latin.
was their custom
it
accuracy, and
if
with us a remnant
who were ever clamant
to
pride themselves
anyone in their hearing
upon
failed to
their
pronounce
such words as antlquus, amicus, fortuna, or imperdtor, with

the accent on the long vowel of the penult, or tdrditas,
trepidits, erirmis, or pdries,
with the accent on the ante-
penult, the offending person was promptly cast without the

pale.
They seemed
amazing blunders
all
such words as
to
be curiously unconscious of their own
they would shorten the long vowels in
sol,
non, vox, nil, lex, lectus^, rex, rectum
they would lengthen tKe short vowel in the accented syllable

of ndta, rdsa, tSga, vdlens, Mne, dSus, hdnus, sdcius, hieme,
they frequently accented the antepenult instead
socletas;
of the penult in viintster, facultas, liMHas, exdllens; and

so with
many
other analogous words.
And
if
they made
use of the phrases bond fide or s^nc qua non, fide was
to rime with " tidy,"
moreover,
still
and sin^ with
"
briny."
occasionally hear of poor "
made
Do we
not,
Smith mine-er
(mindr) being called upon, before he makes another attempt to

' That IB, the participle of ligo.
Lectus, a couch,
and in this case the e is very probably short.
Ib of difieient origin,
PREFACE
VI
struggle through a few lines from the "Mete(r)m<^rphe(r)ses"

(Metamorphdses), to join in singing " duU'sy dome'uni " (dillce
And
domum) ?
with musty
is
"
does not the old-time K.C., engrossed as he

nice-eye pry-us " {nisi priUs) records,
still
get rid of a troublesome case by putting forward a "nolly
prossy-kwy
"
(noM
The remnant
the
"
prosequi)
of this old school pleads for a retention of
old " or " English " pronunciation because, they say,
would save trouble in teaching.
it
Is not the poor British
school-boy worried enough already'
Let him, then, pro-
nounce Latin as he pronounces English.

Obviously, however, the so-called "English" pronunciation
of Latin
is,
in
a measure, a misnomer.
for instance,
If,
we
pronounced fide and sin^ after the English pattern, both

words would be monosyllables
and such words as antiquus,
amicus, casttgo, minister, magtster, libSrtas, facultas, honistas,
mdchina,
excSllens,
would, as
regards
accent,
the
follow
pronunciation of the analogous English words antiquated,

amicable,
The
minister,
castigate,
honesty, machine,
fact
is,
and
many
magistrate,
liberty,
faculty,
excellent, respectively.
of those
who
desire to adhere to the
old or English pronunciation have realised not only
how
knowledge of Latin quantities, but
also
inaccurate
is
their
that, in the adoption of the restored or
Roman
pronunciation,
they are necessarily in danger of being found out.
tomed
'
as they
Accushave been almost from the cradle to the use
The wise schoolmaster does
not, of course, worry his pupils with rules
about the minutiae of Latin pronunciation.

sure that his
own pronunciation
is
accurate.
Nevertheless he does
Then he has
little
make
trouble.
PREFACE
of the inconsistent and purely
artificial
so-called English
to adopt a system in which
pronunciation, they hesitate
demand such scrupulous
quantities
Vll
attention.
Opposition to reform sometimes comes from the most
The Guardian, for instance, regards

the new pronunciation as " harsh, fantastic, and illogical, to
unexpected quarters.
"
the last degree."

for
Harsh
"
Has
The Guardian no ear
musical sounds or sonorous speech
illogical "
Is
seriously
it
"
Fantastic and
suggested that the reformers
are capricious or that the results of their
work are not
substantiated by facts which hardl}' ever admit of question
Then again Professor Saintsbury tells The Times that he

" plumps " for the old pronunciation " because it is more
likely to bring home to an English boy the beauty of Latin
literature."
Assuredly this
is
a thesis that even such a
practised dialectician as Professor Saintsbury would not care

to defend from a public platform.
Yet there are probably
not a few scholars whose keen appreciation of the music of

their Horace
and
their Virgil is so intimately associated
with the pronunciation of a life-time that they cannot bring

themselves to
make a
change.
Could they but
realise
how
greatly the pleasure which they might derive from their
reading
is
diminished by the imperfect rhythm resulting
from a disregard of quantities, they would probably make an

effort to
adopt the more consistent, more
accurate pronunciation which
is
now
logical,
and more
so generally in use.
Unfortunately, a considerable proportion even of those
who now claim
to
use the restored pronunciation
have
adopted the change in a half-hearted and incomplete fashion.
via
PEEFACE
Practically
have done
they
all
new vowel sounds (more

nantal
and
It
u.
is
make use
to
is
or less),
common thing
for correct syllable
length to receive practically no attention at

for the reading of Latin
accent-rhythm
verse,
almost as
is
of the
and sometimes of conso-
all
and
as
the old-fashioned English
much
vogue now as
in
it
was
twenty years ago.

This book
is
not written for schoolmasters
nicht Eulen nach
Athen
for (1) private students
and
tragen.
who
to read Latin correctly
It
is
Wir
wollen
intended principally
desire to learn to pronounce
and (2) those who
acquired pronunciation needs overhauling.
feel
that their
Considerations
of space have, for the most part, resulted in the omission of

discussions concerning the researches which have led to the
adoption of the rules here laid down.

the book embodies comparatively
Of
little.
original research,
It is largely the
outcome of an examination of the researches of others, and

of a checking
and
sifting
which the
of the evidence on
Numerous
results of those researches are based.
authorities
have been consulted and when on any point there has been
any serious disagreement amongst them, the preponderating

weight of opinion has been allowed to decide.
say, this has
Needless to
never involved a mere counting of heads.
Although the main object of the book
to deal with
is
quantity and accent, a preliminary discussion on the various
Latin speech-sounds seems to be indispensable.
method
of giving roughly
to the various Latin vowels

satisfactory, yet
The
usual
approximate English equivalents
and diphthongs
no other method
is
is
by no means
possible
unless the
PREFACE
reader
prepared to take
is
systematically and
IX
up the subject
Dr
thoroughly.
E.
R.
of phonetics
Edwards' has
pointed out to the author the undoubted danger of any
system of mere imitated pronunciation.

particular Latin sound
as a
is
we
say that a
to be uttered in the
same way
If
English sound, do we mean the English

sound as heard in London, or in Devonshire, or in Yorkshire,
particular
or in Scotland, or in America, or
Dr Edwards says, an
a
to result in
will
imitated pronunciation
dialectical pronunciation,
be that we
of Latin as
where?
is
almost certain
and the consequence
have as many dialectical pronunciations
shall
we have
Obviously, as
of English.
Comparatively few English
people ever lose completely every trace of their

ticular dialect,
own
par-
though perhaps few educated men willingly
admit any departure from standard English in their own

speech-sounds.
nised position
uttered.
is
The
bound to
But
in standard English, there
of the
is
a recog-
organs of speech for every sound
slightest departure
from a particular position
result in a different sound.
It is the function
of phonetics to determine the exact positions for
all
sounds.
Merely to listen to a sound and then to try to imitate

of little use, if only because of the
detecting the slighter defects in one's
can be attained only
if
extreme
own
it is
difficulty
speech.
of
Accuracy
the organs of speech are sn trained
that they can be instantly brought into the exact position
Dr Edwards has been the acknowledged

on the subject of phonetics in this conntry. The leading
foreign anthoritiee are Professor Victor of Marburg, Professor Passj of
Paris, and Professor Jespersen of Copenhagen.
'
Since the death of Sweet,
chief aathority
PREFACE
necessary for the production of any particular sound.
marily the whole matter
To
Pri-
organic, not merely acoustical.
deal adequately with such a large subject as phonetics
book of this
in a
is
Yet
out of the question.
size is
it
seems
necessary to touch upon the subject to some slight extent,

if
only to enable the reader to acquire a clear notion of a
diphthong and to understand the inner nature of our slovenly

English habits of speech.
Europe which drawls

English do.
It
is
probably no nation in
speech and
clips its
words as we
how and
therefore necessary to learn
is
where our speech
its
There
is
slovenly and imperfect, for Latin was
almost free from those very imperfections which are so
common
will
Moreover
in English.
it is
hoped that the reader
be tempted to pursue the study of phonetics for
sake.
If,
however, due attention
is
pages of this book,
in the earlier
various Latin sounds
may be
its
own
paid to the hints given

it
is
believed that the
uttered in such a
way
as to
satisfy all scholars except the highly-trained phonetic expert
But the reader must be under no delusion
absolute accuracy
can be attained only from a serious study of phonetic principles.

is
And
unless he already speaks standard English,
hopeless for
him
to expect to attain
it
even approximate
accuracy in Latin.
In touching upon phonetics,

avoided, for the subject
instruction.
"
technicalities
been made even of the vowel "
open
"
and
"
have been
eminently one which requires oral
Little can be learned from books.
therefore, has
the terms
is
all
No
mention,
triangle," or of
close " as applied to vowels.
Ex-
perience shows that a mere book knowledge of such terms
PREFACE
is
likely to
XI
With a
be worse than useless.
care,
little
however, the terms "wide" and "narrow" as applied to

vowels are easily mastered, and as the distinction
is
funda-
mental in Latin they are considered here in some
detail.
One thing
the reader certainly can obtain from a book on
phonetics,
and that
is
puzzling
speech
problems.
Why,
example, in such a word as
circiter
for
on
light
do we tend to pronounce the

alike
many
-ir-
and the
-er exactly
Some
because
little
the
changing.
knowledge of phonetics
pronunciation of
Present-day pronunciation
pronunciation
of,
for instance,
is
English
necessary if only
is
not fixed but
differs greatly
pronunciation of Latin at which
we
are aiming
a standard
is
pronunciation, namely, the Latin spoken by educated

in
the Augustan period.
Latin differed from this
is,
from the
But the
Elizabethan times.
Romans
That pre- and post-Augustaa

of course, true, but that does not
matter.
After a knowledge of the sound values of the various
Latin vowels and consonants has been acquired,
it is
im-
portant to obtain correct notions of syllable length, for this

is
the key to the correct reading of Latin, prose as well as

It was the
verse.
lengthening of short vowels and the
shortening of long vowels
mainly
the result of regarding
accent in Latin as the same as accent in English
many
make
classical scholars to read Latin in such a

it
entirely unlike the Latin of the
that
way
led
as to
Romans.
Hidden quantities have been dealt with

In spite of all that was said at the meeting
in
some
detail.
of the Classical
PREFACE
Xll
Association in 1912', the author
believes
still
it
idle to
is
contend that attention to hidden quantities involves any
The
appreciable additional difficulty in learning Latin.
first
time a learner meets with the word amicus, he sees the long
vowel and knows that he must pronounce
Why,
it long.
then, should there be any difficulty about such a word as
nolUI
There
no need to
is
about hidden quantities

has to do
is
tell
that will
beginner anything
the
come
All he
later on.
to pronounce all long vowels long.
The
rules
about the greater number of hidden quantities are few and

easily mastered,
probably
make
in exceptional cases
under
and though a learner
occasional mistakes
by bringing
will
the rules, his mistakes will be far fewer than they would be
if
he ignored hidden quantities altogether.
After
all,
the
undecided cases of hidden quantity are relatively few.

It is practically certain that the
element of doubt
still
enshrouding some aspects of the question of hidden quantities

tend rapidly to pass away, once scholars attack the
will
question, as Professor Postgate has done, from the point of
view of phonetics.
Hitherto, too
much
reliance has been
placed upon analogical and inferential evidence.

this is very helpful,
to discovering
but
much remains
the precise
to
No doubt
be done in regard
phonetic laws
underlying the
own work in
makes us almost instinctively follow his
guidance, even when his results disagree with those of certain
speech of the Romans.

direction
this
'
For Bubseqaent discussions, see the references given in the foot-note

end of Chapter IX and on p. 111. Professor Buck's contribntion will
at the
,
Prol'essor Postgate's
repay careful reading.
PREFACE
xiii
eminent German scholars who have devoted special attention

to hidden quantities.
Prosody has been only very lightly touched upon, just
sufficiently to put the reader on his guard against imitating
the heavy accent-rhythm characteristic of English poetry.
The
subject
must
is
a large one, and for general details the reader
refer to
the works of recognised grammarians and
metricians.
The
lists
of words
hoped, be useful to
A
pp.
list
commonly mispronounced
many
will, it is
readers.
of useful books for reference will be found on
109111.
In reading the works of
pronunciation of Latin,
it
various authorities
on
the
has been impossible not to feel
considerable surprise at the daring flights which
some
scholars
have made in their researches into the subject. While the

mass of facts accumulated by some investigators, especially
German
many of
investigators, is suggestive of remarkable industry,
the hypotheses which have been formulated on the
strength of those facts are palpably far-fetched and entirely

unacceptable.
It
is
nothing more than
so often forgotten that
imaginary explanation accounting
put an hypothesis to the
for
a group of
is
and
wholly
facts.
In the
it is,
as a rule, easily po.ssible
test of
experiment, and, in the
case of scientific investigation

to
an hypothesis
mentally constructed
case of rival hypotheses, a crucial experiment often definitely
decides between them.
This was the case, for instance,
with the undulatory and emission hypotheses of the nature
PREFACE
XIV
of light.
But
in classical research, rival hypotheses, possibly
both wrong, certainly not both right,
an indefinite period, and,
for
if
may
any sort of claim to distinction, may give
And
opposed schools of thought.

for
new
such schools, when
make the new
facts square
the hypotheses with the
exist side
by side
originated by scholars with

rise to
two entirely
no uncommon thing
it is
facts are against
them, to try to
with the hypotheses rather than
how the
It is curious, too,
facts.
hypotheses of one classical scholar are adopted as facts by

another, a procedure which would be ridiculed in scientific
investigation.
In analogical reasoning, again,
frequently forgotten
not in-
must be
and that in balancing probabilities

judgment there must be an entire elimiSome investigators seem to experience the
weighed, not counted
and
it is
that points of resemblance

;
in forming a
nation of
self.
greatest difficulty in ridding themselves of preconceptions in
favour of a particular theory.

All this applies especially to
made the nature
many
If ancient testimony clashes with their
much
of those
who have
of Latin "accent" a subject of investigation.
the worse for ancient testimony
own
deductions, so
It is surprising to
what dogmatic statements are made, sometimes in

favour of a stress-accent, sometimes in favour of a pitch-
find
accent, although it is notorious that the facts at present
available are altogether insufficient to
admit of a definite
conclusion.
It is
therefore particularly satisfactory to find a wise
reserve and a cultivated caution exercised by certain present-
day British
classical scholars.
Of
these. Professor Postgate,
XV
PREFACE
W. M. Lindsay, rerank.
The latter's Latin
representing Cambridge, and Professor
presenting Oxford, are in the front
Language
is
universally regarded as an authoritative work
of the greatest weight, while the former's
contributions to philology are well

interested in classical research.
numerous scholarly
known
to all
who
aie
Professor Postgate's views
on the whole question of the pronunciation of Latin are so

temperately put forward and carefully reasoned, so
in accordance with all well authenticated facts,
and
strictly
so closely
in keeping with the best traditions of classical scholarship,
that reference to other authorities
is
almost supererogatory.
Professor Postgate was kind enough to read through both
the MS. and the proof-sheets, and his counsel and criticism
have been of the greatest possible value to the author who,

full responsibility for everything which is
however, accepts
not definitely attached to Professor Postgate's name.
Mr
J.
W.
Mackail, too, the author
is
To
deeply indebted for
valuable suggestions in connection with certain points of
doubt and
diflBculty,
more
especially in
Chapter XII.
F.
Auguit 1913.
W. W.
CONTENTS
CHAFTEB
....
1.
Latin and English vowels
2.
Latin and English diphthongs
...
PAOB
1
13
3.
Latin and English consonants
4.
I and
5.
Doubled consonants
6.
Latin syllables
7.
Elision of vowels.
8.
Quantity
9.
Hidden Quantity
50
10.
Accent
59
11.
Gteneral
12.
Prose and Verse.
13.
Words
14.
Exercises
15.
Selected passages,
as consonants
...
26
29
31
Slurring
40
.43
Remarks on Quantity and Accent
Accent and Ictus
frequently mispronounced
Bibliography
73
81
...
90
95
marked
for reading
102
109
CHAPTER
LATIN AND ENGLISH VOWELS

English
1.
to be phonetic
in speaking,
spelling
when
and
it
is
not phonetic.
Spelling
said
is
indicates exactly the sounds
made
for this purpose it is necessary
(6)
That there should be a symbol for each spoken

sound
That each symbol should stand for only one
(c)
That
(a)
sound
in
writing a word no sound should be
omitted
(d)
That no unpronounced symbols should be
English spelling violates
hand,
is
all
Hence there
is
little
Latin, on the other
these rules.
almost perfectly phonetic
difficulty in
used.
so
was Anglo-Saxon.
learning to pronounce
correctly either Latin or Anglo-Saxon.
The foundation
2.
of speech
is
breath expelled by the
lungs and variously modified in the throat and mouth.
voweP
is
merely voiced breath emitted through the
Every time we move the tongue and

new resonance chamber, which moulds the
open mouth channel.

lips
we
create a
voice into a dififerent vowel.

>
But the
possible positions of
In the foimation of nasal vowels, voiced breath flowB through the nose
as well as throngh the mouth.
W.
the tongue and
lips are infinite
it
therefore follows that
number of possible vowel sounds is infinite. No wonder,

then, that we have such a large number of different vowels
the
in different languages.
The difficulty in understanding English vowel sounds

due to the unfortunate names we give to the
symbols a, e, i, o, and u. The names of all five symbols are
The same
really diphthongs, and are therefore misleading.
3.
is
chiefly
symbols in Latin represent pure vowels.
But even
symbols
in Latin each of the five
a,
e,
i,
o,
and u has to represent two sounds, these being usually

distinguished as long and short for instance, we have long
a (written a) and short a (written S,). But these terms long
and short are not wholly adequate, for while there is always
a difference of time there is usually also a further difiference
;
of quality.
4.
To understand these
differences,
learn to isolate a particular sound from
containing
is
necessary to
any given word
it.
Isolate the vowel sound
(a)
To do
it
in the
word
father.
pronounce the whole word, then cut off

syllable -ther, and afterwards the/
Keep the
this, first
the final
mouth open, prolong the vowel sound, and carefully

note, by means of a mirror, the position of the tongue.
Note also the muscular sensation, in order that the
same position may be recognised again.
Now, without in any way changing the position of
the tongue, shorten the vowel sound.
The sound is
exactly the same as before, except as regards the time
taken to pronounce
it.
The
quality of the sound
is
the

same.
however, there
If,
is
the slightest alteration in
the tongue position, or indeed in any part of the mouth

or throat,
some
different vowel
This short a sound

English, though
it is
is
sound
will result.
really extinct
in
modern
very nearly heard in the second
word grandfather, and in the first syllable

of a in the word man is not of the
same quality as the short a in question.
The long a sound in the word father and in the
second syllable of aha is the Latin long a the short a
sound in the second syllable of grandfather and in the
syllable of the
The sound
of aha.
first syllable
short
of aha
is,
as nearly as possible, the Latin
a.
Isolate the short vowel
(6)
sound in the word
fit,
pronouncing the whole word, then cutting ofi" the

Now try to lengthen the sound
and then the /.
first
t,
without allowing
it
to pass into the
sound in the word
The tongue should remain
machine.
quite lax.
If
it
becomes tense or taut, the sound in machine will result,

and this is not only a longer sound as regards time, but
is different
Now
long
in quality.
utter the two sounds, the short
in machine, alternately.
the tongue position, and
in
fit,
and the
Watch the change
in
note the respective muscular
In the case of the short i, the tongue lies

it is thin and wide, and " wide " is
the term used to describe such a sound. In the case
of the long i, the surface of the tongue becomes more
sensations.
easy and relaxed
The
is a marked feeling of tenseness.
bunched up and narrow, and " narrow " is the
convex and there
tongue
term used to describe such a sound.
is
12

Isolate the long vowels in the words fate, note,
(c)
and
Note the tense
hrute.
The vowels
each case.
feeling of the tongue in
are narrow.
Isolate the short vowels in the
Note the easy relaxed
full.
The vowels
words sped,
not,
and
feeling of the tongue.
are wide.
and
There is a diflference of length only. There is
note.
no difference of quality. With these, compare the two
a sounds and the two i sounds already described.
Isolate the long narrow o sound in the word
(e)
As the sound ceases, an " easing
note, and prolong it.
Isolate the o sounds in the words ohey
(d)
down
"
The pure
The unconscious
sound
is
alteration
alteration of sound,
by an
be de-
thus not cut off promptly.

of
produces an
position
and the pure
sound
is
followed
The English
easily perceptible secondary sound'
long
may
of the tense feeling of the tongue
tected.
therefore really a diphthong, though the second
is
element is slight. The secondary and final sound is a

u sound, and is very distinct when it comes at the end
of a word as in no.
It is scarcely heard in such a word
as noble.
Cut ofif this secondary sound and we have
the pure vowel sound heard in the German word Sohn.
This
is
the Latin long
(/)
Note
its
final i
o.
Isolate the long vowel
sound.
in such a
sound in the word
fate.
diphthongal nature, due to a secondary and
The secondary sound is scarcely heard

Cut ofif the secondary sound,
word as nation.
and we have the pure vowel sound heard

word fehlen. This is the Latin long e.
1
Cf.
13 and 14 in the next chapter.
in
the
German

{g)
Note
its
word machine.
Isolate the long i sound in the
diphthongal nature, due to a secondary and
Cut off the secondary sound and we

y sound.
have the pure vowel sound heard in the German ihn
or French livre.
This is the Latin long i.
final
(h)
Note
its
final
the vowel sound in the word brute.
Isolate
diphthongal nature due to the secondary and
Cut
sound.
off
the secondary sound and we
have the pure vowel sound heard in German du or

French rouge. This is the Latin long u.
6.
It will
to the symbols
now he
a
clear that the English
(as in fate), e (as in
(as in note), are not
names we give
seen), and
machine or
pure vowels; they are diphthongs.
The English names we give to the symbols i (as in pine),

and u (as in duke), are referred to in 14.
Even the sound of a in father is possibly slightly
diphthongal, but the secondary and
as almost to be imperceptible,
and
final
is in
sound
is
so slight
consequence diBcult
to analyse.
6.
The Latin
They
pronounce.
short vowels, d,
are
heard
S, %,
6 and
-it,
respectively in
are easy to
the English
words grandfather (second syllable)\ sped, fit, not, and pull.

There is some difference of opinion about Latin short o.
Some authorities consider it to be merely the shortened
form of the long o (as in obey or democrat), but the weight
of opinion is in favour of the o sound in not.
7. The only new vowel sound to be learned in Latin is

the y found in certain words borrowed from the Greek. It'
But
Bee last paragraph of 4 (a).
6
is
French
really the
u, or
German
The sound
ii.
from a combination of the tongue articulation of the
results
i
sound
and the lip-rounding of the u sound.
sound in the word brute, and

very carefully observe the rounding of the lips.
(a)
Isolate the vowel
(b)
Now
isolate
Note
machine.
the long
sound in the word
the exact position of the tongue
retain that position.
Now
round the
lips
and
as for the
u sound in brute, and again utter the long i sound.

The vowel sound in French nise or German Oiite
This
results.
Now
(c)
in the
word
The
Latin y
Latin y
Latin long y
is
The vowel sound
German Siinde
in
very nearly, but not quite, Latin short
quality of the Latin p
the diflference
is
(y).
do the same with the short vowel sound

fit.
This
results.
(p).
is
is
is
the same as the
difiference of length only.
almost exactly heard in French
cultiver,
where we have the same " narrow," " tense " sound, but
Perhaps the sound is best
shorter, as in French mse.
obtained by isolating the u in French ruê and then
shortening
it,
be kept tense.
but in the shortening the tongue must
As
in
French
rttse
and
Latin y and p (really Greek v and v)

the sound is the same.
cultiver, so
;
with
the quality of
8. In certain Latin words the spelling varies between

and u, for instance, mazimus and maxumus, optimus
and optumuB, and there is little doubt that in these cases
the actual vowel sound heard was intermediate between i
and u, and practically equivalent to Latin p, as in German
Siinde see above. It was a lax, not a tense sound, and
i
therefore not quite the same as Greek v.

The Latin p
sound, rendered lax, should be adopted for such words.
9. Latin short vowels, when final, are very seldom

pronounced correctly. For final short a (as in viUS.) the
tendency is to substitute the vague and indefinite sound
heard at the end of the English word china. For final ?

and i the tendency is to substitute the short sound heard
at the end of the English words lady, sorry; thus, poBs
is often pronounced as possy, and nisi as nissy.
For final
.
is
8,
often substituted; thus
(or even modo).
mddo
These are very bad
The best way
is
pronounced mddd
faults.
to obtain the correct pronunciation
is first
to pronounce the
first
word with an added t^ at the end. Thus,

pronounce the imaginary words villdt, possU, nisit,
mdddt, distinctly;
then cut off the
and preserve the

Except
for the different position of the break at the end of the
word, pared tihi and parcet ibi, rgggri testis and rggerdt
MAb are pronounced exactly alike. In such a word as
t,
accurate pronunciation of the final short vowel.
hiiii it is best
form hitdmdt
first
to pronounce correctly the imaginary
then omit the two
t's
and
carefully preserve
the vowel sounds.
The following
10.
table includes the twelve Latin vowel
Words from
sounds already discussed.
German
English, French, and
are given to illustrate the sounds, but the words in
brackets give the sounds only approximately.

are, of course, to
'
One
the letter
"
of the best letteiB.

r.
Some
For the reason of
This table
is
The
references
standard English, French, and German".

would be misleading, for instance
any text-book on phonetics.
can be made for those unfamiliar with
letters
this, see
as accurate as
it
Latin
vowel
LATIN AND ENGLISH DIPHTHONGS
CHAPTER
II

12.
Latin diphthongs are seldom pronounced correctly,
but the pronunciation

a diphthong
13.
is
is easily
mastered, once the nature of
understood.
glide
is
a transitional sound, generally used un-
consciously, produced during the transition from one
sound
Pronounce the word key: here we have the

initial consonant (a back position) sound, and the final vowel
(a front position) sound.
The glide is the sound produced
to the other.
With a
in passing from one to the other.
the
little practice,
muscular movement in the production of the glide may be

detected.
To the student
ance, but
it will
of phonetics, glides are of great import-
suffice
here
if
the reader can distinguish
between the initial and the

diphthongal vowel sound.
clearly
14.
sound
We
final
element
of
have seen that when, in pronouncing the vowel
in such a
word as
no, the
sound
unconscious alteration of position, a
ceases, there is
final
an
and secondary
sound being produced.

But there may be a conscious change from one vowel
In such a case, the initial and final
position to another.
vowels, with the vowel glide between them, constitute a
diphthong.
Each
" syllabicness
"
of the two vowels virtually loses its
own
the vocal organs shift rapidly from one
position to the other,
and there
is
combination of sound.
10
though one of the vowels takes the chief
stress.
practice with a few English diphthongs will
make
this clear.
little
and pronounce in succession the Latin

a and i, first slowly, and then more
rapidly, until the two coalesce and form the diphthong
ai.
Note how the a is stressed, and how weak the i in
the combination becomes.
The combined sound is the
Isolate
(a)
short vowel sounds
name we
give to the English letter
sound heard in the word pine.
i,
that
is
the vowel
(Do not be misled by
the combination ai in such a word as maid, which is

quite a different sound.)
Clearly, then, the
name we
give to the English letter
a diphthong.
i is
(6)
Now
resolve this diphthong, that is the vowel
sound heard in the word pine, into

the diphthong
first in
its
elements, uttering
the ordinary way, then more and
more slowly until the separate a and i sounds are

(Note if the element a be lengthened to
the long a in the word father, and the diphthong be
reconstituted, the cockney pronunciation of the i sound
detected.
in
pine will result.)

(c)
Resolve the diphthong ou in the English word
loud into
its
elements.
They are, respectively, Latin a

Latin a be shortened to Latin &,
(Note if the
a,
and the diphthong be reconstituted, the cockney pronunciation of ou will result for instance, in the word
and
house.)
(d)
(Cf.
f.
6.)
Resolve the diphthong
moist into
and
16
its
elements
they
oi in
the English word
are, respectively,
Latin 8
11
(e)
Resolve the diphthong u in the English word
duke into its elements they are, respectively, Latin 1
and u. This diphthong is the name we give to the
;
English letter
u.
16. Not one of these four diphthongs occurs in Latin,

but the reader ought now to be able to pronounce those
that do so occur. They are six in number. Three of them,
viz., ae, au, oe, are common
the other three, viz., el, eu, ui,
;
are rare.
16.
The
Latin diphthongs have no exact equivaand their pronunciation must be learnt

sounding each Latin vowel separately, and then
six
lents in English,
by first
running them together.
If attention be paid to the sugges-
tions in the preceding section, there should be no difficulty
in effecting the fusion accurately.
The
following remarks
(a)
helpful
ae (as in taedae) may be pronounced as

when pronounced broadly. It is nearly like
Lat.
in Isaiah^
Fr. travail gives the
Ger. a.
the
may be
sound
fairly accurately if
vanishing sound of the liquid be omitted.
final
The sound
is
never like ai in Eng. maid, or
ee
in
Eng. feed.
(b)
Lat.
au
(as in
laudo)
is
the ou of Eng. house
broadly pronounced, though not quite so broad as in
the cockney pronunciation.
Ger. Hatis (containing
more
of the w sound than Eng. house) gives the Latin sound

exactly.
But
it is
not certain that Lat.
au
is
-(-
in
' Old Latin at was, as Professor Fostgate points oat, certainly pronounced as in Isaiah, but when ai was changed to ae there was probably
a change in the sound of the second element from i to e. So with oe (at).
12
all
cases.
It
Latin sound
may sometimes have been + u. The

au is never like au in Eng. fravd or in
fi,
Fr. pauvre.
Lat. oe (as in foedus)
(c)
Eng.
boil
tion.
It is never like ee in
something like
is
the du in Ger. Hauser
also
is
ot in
an approxima-
Eng. feed.
something like ei in Eng. rein or ey

in grey.
It is a diphthong in hei and Pompei (voc).
But the combination ei is far more common in separate
(d)
Lat. ei
syllables,
e.g.
is
diSi, fidei.
deinde is not quite

the word is always
free
The nature
disyllabic,
and there appears to be
no good reason to think that
nounced in
it
was differently pro-
prose.
Lat.
(e)
of the ei in
In classical verse
from doubt.
eu
(as in neuter) is very nearly like the
ou in the cockney pronunciation of Eng.

never like eu in Eng. feud.
Lat. heu
like the cockney how.
(It is perhaps a
if
eu
in Lat.
trisyllabic at
(/)
neuter
is
a diphthong.
very
little
It
It
hcruse.
is
is
much
doubtful
was certainly
one period'.)
Lat. ui (as in huic)
is
very nearly like Fr. oui
But ui almost always forms two syllables

as in monui, fdit. As a diphthong it is a little difficult
to pronounce, the two vowels not coalescing very easily.
The Eng. word fluid gives only a rough approximation,
(from ouir).
the two vowels fusing only in a very slight degree.

'
Professor Postgate has pointed out that the proniuiciation of neater is

an unsolved problem. " It is difficult to find a passage in Classical
really
verse where the first syllable mutt be long.
On
the other hand, nifttor
cannot be proved, as there was a form necnter, a form which Classical

poets (Lucretius, Seneca, Martial) used if they wanted a trisyllable."
LATIN AND ENGLISH CONSONANTS
The pronunciation
In qui the
fused.
word
strikes
thing like
disyllable
is
a consonant, and qvi

i
In cui the ui
upon the ear as
'coo-y,'
qui are often con= kw;

being clearly pronounced
of cul and
thus qui = kwee, the vowel

as a simple vowel.
13
a diphthong
is
the
the spelling were some-
if
but anything of the nature of a
must be avoided.
CHAPTER
III

17. Consonants are the result of audible friction or
stopping of the breath in some part of the mouth or throat.
One convenient
basis of classification' of consonants de-
pends upon the degree of obstruction of the mouth channel.
From
this point of view, consonants are either mutes, or
contitmants.
If the
mouth channel is
when
" explosion " takes place
closed or entirely stopped, an
the stoppage
is
two
" labials,"
b,
two
''
dentals," d,
The
number
removed.
explosive consonants, also called mutes, are six in
and two so-called
" gutturals,"
If the
g (as in go), k.
mouth channel is not quite
produces a rubbing or
" fricative "
closed, the obstruction
sound.
The
fricative
consonants are also called continuants, since the sound can
be kept up by merely continuing the breath.

continuant
1
is,
The term
however, rather loosely used.
Consonants are classified on diSeient bases, not always logical in

Only sncb points are touched on here as are necessary for
character.
dealing with difficalties in Latin.
14
Certain fricative consonants are

of these,
and
known
as liquids.
consonant combinations.
into certain Latin
Two
for special notice, as they enter largely
r, call
Liquids are
formed with so wide a passage, and with so little friction,

that they must be regarded as closely akin to the vowels.
The consonant w
is
also closely akin to the vowels, as will
be seen later on.
Both
I,
r and
"
are sometimes described as " vowel-like
in character.
18.
One
of the
to take the shortest
This often results in combinations without
juxtaposition.
any
" glide " at all.
Combinations of mutes and the voweland w (for instance, br, pr, dr, tr, gr, hr,
hw) are glideless in both English and Latin,
like consonants r,
hi,
most general principles in speech is

in immediate
way between two sounds
pi,
cl,
gl,
I,
immediate fusion taking place. For all practical purposes,

such combinations are simple sounds there is one and only
one effort of the voice made in producing them. This may
be tested in such words as hray, pray, dray, tray, gray, cray
{=krayy,play, clay {=klay), blow, glow, quite (=hmte). Sec.
;
19.
It is particularly necessary not to confuse other
combinations, especially
sc,
sp,
and
with the glideless
st,
Such confusion
combinations in the preceding section.

the cause of one very
common
of Latin, as will presently be seen.
experienced
in
passing from
is
fault in the pronunciation
the s to
distinct effort is
the
mute.
The
pronunciation of such pairs of words as car and scar, crawl

Note that in English e ia a redundant letter, for it has the sonnd of
Other redundant letters are q, for qu = kw; and x, for x has
either the sound of ks as in extra, or gz as in exact.
'
either s or k.
15
and scrawl, pin and spin, play and splay, top and
and strain, should be carefully compared.
stop, train
The Latin consonants may now be touched upon
As in English.
Lat. urbB, absorbeo, as
ao.
b.
21
bs.
in Eng. caps
seriatim.
never the bz
sound in Eng. cabs. Lat. s never takes the Eng. z

sound, and to preserve the b sound in such a word as
urbs, the
32.
bt.
It is a case of
naturally passes into p.
the law of least
Lat. obtineo,
the
tion, as in
effort'.
as in Eng. kept.
last,
In this combina-
the change from
place almost unconsciously.
Even
to
takes
in the English
word obtain some effort is necessary to prevent the b

from becoming a p, and a still greater effort is necessary to preserve the b and the s in such a word as
is
correctly sounded, the s tends to
become a ^ if the
to become a p'^.
s is correctly sounded, the b tends
absorb ; if the b
;
It
may
be noted that apsens and optineo are good
spellings in classical Latin.
23.
c.
Lat.
cftno, c6clni, condlclo, as in
never the s sound in
circle.
Hence
Eng.
cot, kite
kekinee, condikio
(never condishio).
24.
Lat. pulcher, Bacchus, as Eng. k followed by

an aspirate; never as in church or chivalry. The
ch.
ba the soTmd began as a 5 and finished as a. p.

These tendencies are perfectly natoral and are easily understood, once
the relations amongst the different consonants are understood. Consult
any text-book on phonetics.
1
'
It is possible that in Lat.
16
by pronouncing the
correct sound can be obtained
words deck-hand or fork-handle in such a way that

the
mute k comes into the second syllable, e.g., forThe sound is easy to an Irishman. It is
khandle.
better to omit the aspirate altogether, as was done in

old
25.
and Popular Latin,
As in English^
d.
if
But
there
is
before a
any
difficulty.
word beginning with
a consonant, d final often becomes a
haud, apud.
So Eng.
tidbit
became
^,
as in sed,
titbit.
Cf
Ger.
unt for und.
As
26.
f.
27.
g.
in English'.
Lat. gaudeo, genus, as in Eng. got, get; never as
j in jet or g in gin.
After a vowel, as
regnum, cognomen, gn was

reng-num, c5ngn5men. But on l^his point there is doubt, and gn
is best pronounced as in English (e.g. reg-num).
28. gn.
in
possibly pronounced ng-n, thus
29. gu.
In
u may be a vowel or a
an ordinary vowel in arguo, in
(e.g. frigui), and in adjectives in
the combination gu,
consonant.
It
is
perfects in
-gili
-guiiB
ambiguus);
(e.g.
in all these cases,
gu has
the sound of goo in Eng. good.
After
See
30.
h.
the
gu
is
generally consonantal.
As in English, but more slightly the breathing

must only just be heard. Lat. h is sometimes a mere
Nearly.
In
all probability
tongue touching the teeth.

2
in
38.
But with more "breath."
the Latin Bound was pronounced with the
17
sign of hiatus,
i.e. it is used to make clear to the eye

that the vowels are to be pronounced as two syllables
rather than as a diphthong, e.g. in ahSnus (another
spelling of aenus).
Lat.
h never
prevents "slurring"
or the shortening of vowels before other vowels (see

73, &c.),
and
it
often admits contraction, as nil for
nihil (or nihil).
h never makes
Lat.
31.
32.
33.
34.
35.
position.
See
67.
As English.
As English.
m. As
English, except when
See
As English.
no. Lat.
Here nc = ng + k; incipit = ingk.
in
in
1.
in
n.
final.
52.
in
incipit.
kipit.
As
in English, n before
"
guttural
"
is
so
affected as to leave its proper sound incomplete (the
tongue not touching the roof

cord
36.
nf.
mouth), while
of the
tends to draw the guttural into
itself.
So Eng.
it
con-
= cong-kord.
In
Latin words, n in the combination nf had

lost its consonantal character altogether,
probably
leaving the preceding vowel nasalised and long.

it
is
combination as in English,
e.g.,
the preceding vowel in Latin

of this there
is
for instance, is
Some
Cf.
81
But
Eng. bon-fire.
is
invariably long
no doubt whatever.
pronounced
clasBical scholars prefer to
the preceding long vowel.
W.
But
perhaps best' to pronounce the consonantal
Lat. conflcio,
cone-fikio.
omit the n altogether, and to uasaliee
(1).
AND ENGLISH CONSONANTS
LATIN
18
37. ng.
Lat.
Lat.
38. ngu.
ng
is
In
+ g; congero =
= ang-ger (see 35 above).
Here ng = ng
congero.
cong-gero. So Eng. anger
never like ng in Eng. singer or hanger.
Latin,
before another vowel
is
generally
pronounced as Eng. w in wine.

Hence from 37 it follows that
(See 55, &c.)
anguis is pronounced ang-g^ris, and lingua, lingconsonantal, and
gwa;
so with
of which
all
is
sanguis, pinguis, unguis, languor,

of course,
are,
anguish (= ang-gwish).
the
with
is
a vowel, and
37.
Cf Eng.
dissyllabic.
In such a word as longUB,

is
pronounced in accordance
Cf. also 29.
39. nq. Lat. inquam.

Here nq = ng + q; inquam =
ing-quam. So Eng. relinquish = reling-quish. Cf
40.
ns.
35 and 37.
In
Latin words, n in
probably
lost
the
combination ns had
consonantal character altogether,
its
leaving the preceding vowel nasalised and long (cf

36).
It is perhaps best' to
pronounce the consonantal
combination as in English,
e.g.
Eng. in-sight.
But
the preceding vowel in Latin, as in the case of nf,

invariably long.
nounced
41.
42.
Lat. pinso, for
instance,
is
is
pro-
peen-so.
As in English.
ph. Lat. phalerae, Phoebus, as Eng. p followed
p.
aspirate
never as in Philip or elephant.
The
by an
correct
sound can be obtained by pronouncing the words

'
Some
classical scholars prefer to
the preceding long vowel.
Cf.
omit the n altogether, and to nasalise
36 and 81
(1).
19
strap-hanger or mop-handle in such a way that the
mute p comes
phanger^.
into
omit the h altogether.
43, qu.
Lat.
second
the
If there is
any
syllable,
stra-
Cf. 24.
aquila as in Eng. queen.
In Latin, as in
English,
qu must
not be regarded as the consonant
is
e.g.
difficulty, it is better to
qu = Aw
q followed
always followed by u, and Lat.
by the vowel u, but as a consonant group kw, always

followed by some vowel.
The u of qu must always
be treated as a consonant, just as the second element
of any other consonant group (e.g. cr) is treated (see
18); qu cannot, of course, form a syllable by itself,
since it
qu
is
is
Lat.
44.
r.
In
nothing more than a iabialised
k.
In verse,
always regarded as a single consonant.
qu
is
never sounded as
sounding
the hard palate,
r,
is
in Fr. quatre.
the tongue, after almost touching
made -to
vibrate towards the upper
gums. Hence
Except in the North it is never really heard in
England as a consonant, unless it is followed by a
vowel in the same or in the next word.
In the
South of England the words father and farther are
pronounced exactly alike, though many people exr has been called the trilled consonant.
perience difficulty in realising
this.
Pronounce the words hear them, and note that

the r is not sounded there is a mere voice murmur
;
only.
Now
1
pronounce hear in hear them exactly as in
This suggested pionunciation must not be confused with the pronun-
ciation sometimes given now-a-days to "strap-hanger," viz.,
" stra-f anger."
22
20
hear
The consonantal sound
it.
of the r can
now be
detected.
In Latin, the r
and Scotch.
is
always
proper rolling of the r
is
unaccented
to slur all
-er,
-ir,
rule
no
-ur,
The English tendency

Thus we pronounce
final
is
?)
usually given to ver, cur, and vir
pronounced wehr, koor, and wir).
syllables
among
and as a
is
Arbor
The mispronunciation
not ahba but ahrbor.
is
distinction,
made between Lat. mater,

and (a)mata. The same vowel sound
(a)inatur,
(correctly
The
finals.
and -a without
difference
(and no
French
extremely important, in
finals as well as elsewhere.

is
as in
trilled,
Lat. rariis, dator, as in Fr. dire.
ending in r
is
very
of
common even
classical scholars.
Note that such a word as audirem should be

pronounced, save for accent, as audi rem. An indistinct
the
rh.
Lat.
is
vowel sound
is
often wrongly inserted after
(as in Eng. dearest)
sound
45.
is
in other words the pure

wrongly converted into a diphthong.
;
rhetor as in Fr. theatre.
(rh
the same sound as r except that
= Gr.
p.)
It
it is voiceless, i.e.,
pronounced without the vocal chords vibrating.

Eng. initial r, as in rub, but
there is no exact equivalent in English.
it is
It is approximately the
46.
B.
Lat. SUB, rosS,, accuBo, as in Eng. hisB, hist; never
as in Eng. has (haz).

reez),
mdns
rosa as
>
But
Pronounce res as race (not
as mohiiss (not monnz),
rossd^,
pars as parss.
miser
Few
as misser^,
letters are
not, of course, in imitation of a doubled consonant.
See Chap.
so
t.
mispronounced
frequently
= sk
Bc
Lat.
as
21
Note
s.
that
pronounce scientia as sMSntidi, scite as
skite, &c.
47.
In Latin, the u in su is generally a consonant

(= Eng. w) in suadeo, suasi, suasum, suesco, BuSvi,
su.
suetum, Buavis, suavitas.
On
swahwis, &c.)
suadeo
in
or edition.
Lat. editio
th.
in
in
su
t is
Eng. table'.
is
Never
as in Eng. nation
never edishio.
Lat. thalSmils, Cethegus, as Eng.
The
followed by an
correct
sound
be obtained by pronouncing the words gift-horse
or hot-house in such a
way that the mute
into the second syllable,
any
The sound
invariable.
aspirate, never as in thin or then.
may
being followed
see 59, 61).
Always as
of Lat.
49.
The u
so with the other words.
by a consonant
t.
a vowel in sui, suus, sudrum, suere, &c. (and of
course in bus, super, sutor, &c., the
48.
swahdeo,
when the
a vowel, the word contains four syllables
is
(su-a-de-o)
is
(Pronounce
the rare occasions
e.g.,
gif-thorse.
A
Note: ch, ph, th were introduced
difficulty the
century
Greek
B.C.
comes
If there
is
should be dropped altogether.

in the first
to represent, in borrowed words, the
aspirates,
which had previously been repree.g., teatrum, later
sented by the simple mutes,
theatrum.
They came
genuine Latin words
to be used also in a few
''.
All three sounds are easy to Irishmen.
Nearly.
more "dental" than in English. Fr.

Eng. table.
pronunciation as c, p, t, was probably not uncommon.
Like
gives the Lat. t

2
But
their
d, the t is
more
correctly than
table
22
50.
Lat. sSxiim,
Eng. extra, never the gz

= eksulto (not egzultai).
exulto
sound in example
Note that x is a "double" consonant, and is a mere
representative of cs or ks. Regsi (from rego) first
X.
fix&lto, as in
;
became
Lat.
z.
recsi, then rexi.
gSzft, zonS,, as in Eng. adze, not like z in zeal.

was certainly a double consonant, for it made
position," and this the Eng. z (which is only a voiced
Lat. z
"
sy could not
do.
and voiceless consonants is important

and lengthen the consonant /in the word^t (-ffff-< not the name
ef), and then the corresponding consonant v in the word liver (-vvm>-, not
the name vee).
Note that while /is articulated in one place only, being the
result of the friction of the outgoing air between the lower lip and upper
1
The
distinction between voiced
Isolate
teeth, V is articulated in
and
two places,
in
the throat as well as between lip
teeth.
Place the
first
The vibration which

two fingers on "Adam's apple."
felt, but this is not so in the case
produces the effect of voice in v can be

of/.
Breathe in the ordinary way, then bring the lower

together
lip
an / is produced.
Make a
voice
and teeth together a d is produced.

Do the same with the pair s and
murmur,
and upper teeth

and bring the
lip
as in err,
breath,
ssis, zzzz, sszz, zzss,
until
z,
first
separately
the distinction
and then in one

and thoroughly
is felt
under command.
Do
the same with the th in thin and then.

Pronounce kaa and gaa. Now cut off the vowels and try to sound the
consonants alone. In the case of the g an audible "guggle" is made, but
no sound is made in the case of k, although there is a feeling of muscular
tension in the tongue.
The
distinction ought
are voiced
y,
p,
t,
and
and zh are voiced
are voiced.
A;
now
Of the six " mutes," b, d, and g

Of the continuants, v, w, z, tA(en), j,
The liquids
ch, and sh are voiceless.
to be clear.
are voiceless.
/, xvh, s, th(in),
53.
23
The pronunciation of
final.
final is not free
from doubt, but it was certainly much more weakly
sounded than at the beginning or in the middle of a
word.
In some way or other
final
was greatly
reduced, probably through the lips not being closed

to
pronounce
Possibly
it.
The
of whispered m.
was some adumbration
it
facts
we
are quite sure about
are
(1)
of its
Final
did not prevent the preceding vowel
own word, and the
vowel of the following
initial
word, from being reckoned as one syllable, precisely

as would have been the case
Final
(2)
if
no
m had
intervened.
followed by a word beginning with
a consonant invariably
made the
which
syllable
it
terminated long.
is some difference of opinion as to the
method of pronunciation, but the weight of
There
correct
authority
is
in favour of the following rules:
If the next word begins with a vowel (or h),
(1)
drop the m, nasalise the vowel before

this vowel
diflBcult,
gether.)
to the following one.
drop the
and
vi
its
it,
and run on
(If this is too
preceding vowel alto-
Thus
fluctum accipit
is
pronounced flucVaccipit;
alternatively fluct' accipit.
bonam
addit
is
pronounced horfiaddit
alterna-
tively bon'addit.
The pronunciation need occasion little difficulty in

The m must be dropped in any case, and
practice.
if we simply touch lightly on the vowel remaining
24
(which
initial
is always short), and pass on rapidly to the

vowel of the following word, sufficient nasal-
isation will be
not at
is
all
produced almost unconsciously.
necessary to
It
effort to
Neither should the two vowels

The first should only just be
nasalise the sound.
form a diphthong.
heard
make a marked
the second should be distinctly pronounced.
Thus the above pairs of words may be written

flucVaccipit and bonâddit.
If the next word begins with a consonant,
was of opinion that while, as before, the m
should be dropped, its effect should be preserved
by pronouncing the initial consonant of the next
word as if it were doubled, keeping the vowel before
the m short, but running it on to the doubled
(2)
Ellis
Thus
consonant.
infandum regina is pronounced infandurregina

jann6x(not jd.nnox)
jam nox
tantum fata
tantuffata
quum
quuppedes
,,
spargaffl6res
p6des
spdrgam
In
all
flores
cases the accent remains as
if
the words were
separately pronounced.
But
at the present time the weight of opinion is
against Ellis, as he
favour
of a
is
The
the language.
more
"
following consonant.
not supported by the facts of
best authorities are strongly in
organic " assimilation
Hence
to
the
the following subsidiary
rules
(o)
Before
n,
t,
d,
s,
and perhaps /, j
(i
con-
and
sonant),
{u
consonant),
25
changes to n.
Hence
jam nox
is
m^nsam tenet
6quum divina
pronounced jdnnox
tdntum sperabam
spdfgam fl6re8
claiiim jdcit
mensantenet
equundivina
tdntunsperabam
sparganflores
claliinjdcit
mtisam vidit
mtisanvidit
Before
(/3)
c (or q)
and
g,
changes to ng.
Hence
mensam
graliem
is
pronounced mensanggradem
cdntum consiiltor
uterum-que
,,
cantungconsiiltor
uterting-que
quamquam
qudngquam
(7)
m remains m.
Before m, b and p,
Hence
mdgnam mavortis is pronounced m^gnammav6rtis

quidem bonitas
quorum pars
(8)
Before
quidembdnitas
qu6rumpdrs
and
r,
is
completely assimilated.
Hence
jam
Iab6ntibus
is
infdndum regina
Ellis favours
(consonantal
pronounced jallab^ntibus
dropping the
i)
and
mtisam
vidit
altogether before j
v (consonantal u),
ing the vowel (preceding the

claliim jdcit
infd,ndurregma
m)
long.
and pronounc-
Thus
would be pronounced clauijdcit
musavidit.
26
This
may
AND O AS CONSONANTS
be regarded as an acceptable alternative
to (a) above.
been suggested that, if the last word

of a Latin sentence ends in m, the m should be
sounded very lightly, as in the current pronunciation
of Cheltenham, Birmingham (Chelten'm, Birming'm),
But this is certainly wrong, for it is always
&c.
It has
(3)
should
weak, not the vowel. The
be totally inaudible, but the preceding vowel is best
sounded long, if only to indicate the excision of m.
the
that
is
The following English consonantal sounds do not

occur in Latin ch in church j in judge sh in shall, sure,
patient th in thin th in then wh (breathed w) in where
(in Scotland); yh (breathed yy in human; zh in pleasure;
53.
z in zealous, rose.
CHAPTEK IV
I
54.
AND
AS CONSONANTS
In sounding English
touches the hard palate
j,
the front of the tongue
in sounding
hard palate but does not quite touch
it.
ship between Eng. j and Ger. j (which
equivalent to Eng. y)
is
it
comes near the
The
close relation-
is
approximately
thus easily understood.
element of the diphthong u is alone considered in this

That element does not appear in Lat. long u. In Latin, hu
would be pronounced hoo.
>
The
example.
first
In sounding
AND U AS CONSONANTS
the tongue
position as in sounding
(ee).
is
very
27
much
This explains
in the
same
why
tends
to become y before a vowel (e.g., onion is pronounced onyon).

In such words as yet, yonder, y is virtually a consonant if
however in these words i be substituted for y, and the words
;
ionder) be rapidly pronounced, the results are practically,

not identically, the same as before.
{iet,
if
(The difference between i (= ee) and y is that in y the

tongue is put in the i position and immediately withdrawn.
This makes practically an action and so a consonant, as
opposed to a position, i.e., a vowel.)
55. The letter v is merely another form of u and was

used either as a vowel or as a consonant. The letter
is
really a double v, though it is called a double u.
In such words as win and twice,
is virtually a con-
sonant;
if
however
in these
words
{=00, as in food) be
substituted for w, and the words (uin, twice) be rapidly
pronounced, the results are practically the same as before.
56.
Thus both
and
u when pronounced
rapidly before
other vowels become virtually consonants, with a distinctive
sound of their own.

distinction in sound,
in
j
The Romans, though aware of this

made no distinction in writing. But
modern times printers have adopted the
(i.e. i
with a distinctive
tail) for
practice of using
the consonantal sound of
sound y) and of using v (a mere variant of u)

for the consonantal sound of u (i.e. for the sound represented
by Eng. w).
The use of these letters j and v cannot be justified, and
(i.e.
for the
as regards pronunciation
is
very misleading.
28
AND D AS CONSONANTS
57.
Lat.
(2)
Though
Ger.
consonantal
i is
When
When
(1)
and followed by a vowel.
initial
it
occurs between vowels.
often printed
Thus
58.
pronounced
or
yS,klo
iScIo (jXcio)
is
maior (major)
ma-yor
ius (jus)
ytiss
aio
a-yo
a-ye-bam.
aiebam
In certain words borrowed from the Greek,
before a vowel
is
a true vowel,
borrowed Greek words,
two other vowels,
But usually
by a vowel, the
pronounced as Eng.
it is
_;',
j.
or e and
is
in
iambus.
as in Agla-i-a, Zia-i-us.
in the
i
e.g.,
a true vowel even
i is
initial i
In other
when between
So in Ga-i-us.
combinations ai and ei when followed
both makes a diphthong with the preceding
consonantal to the vowel following.
Thus
Staientts
= Stai-ye-niis
= Pom-pei-ytis
Porapei&s
MaiGs
and
in
59.
Consonantal
is
never used as a
vowel.
Though
= Mai-yiis.
never found before a consonant
is
final
often printed
wine, or perhaps
still
it
occurs invariably before a
pronounced as Eng. vr
more nearly like Fr. ou in oui or
it is
ouest.
Thus
u6t6 (vSto)
is
ciuis (civis)
Smault
(ftmavlt)
pronounced w6-to
kee-wiss
a-ma-wit.
DOUBLED CONSONANTS
60.
29
It should be clearly understood that the letters j
and V are mere arbitrary signs used by the printer. The

Classical Association recommends that the use of j shall be
entirely discontinued, and that v shall be used in books for
beginners only.
61
Occasionally consonantal
and consonantal u may
be preceded by a consonant in the same
syllable.
In, for
word suauls (suavis) the consonantal u is

the word is a dissyllable, and is pronounced
instance, the
preceded by 8
swd-wis. (See also
;
Note
47 ^)
In this book consonantal
and u are generally
printed j and v respectively, though in the exercises at

the end of the book they are shown by difference of type
within the same word,
e.g. iacio
or iacio (jacio)
The symbols_/and
or cinis (civis).
with considerable reluctance,
ciuis
have been retained
for there is
no justification
But no doubt words containing them
for their use.
will
be more familiar to the eye of the ordinary reader, and

they have been preserved. It is however
hoped that both symbols will soon disappear from
for this reason
to be
all
Latin texts.
CHAPTER V
DOUBLED CONSONANTS
62.
is
The
correct pronunciation of doubled consonants
of great importance, and
for the
proper rendering of
metre and rhythm, of fundamental importance.

'
Madvig
Bi>eUs svadeo, qvis, etc., in his Latin
Grammar.
DOUBLED CONSONANTS
30
To pronounce
first
a doubled consonant correctly, hold the
element until ready to pronounce the second.
duplication should be regarded as the

of a single consonant.
of expiration,"
"
There should be
"no relaxation
The
energetic utterance
"
"
no discontinuity
of the organs";
and "no puff
"
should intervene between the two

parts of a doubled consonant, which should more resemble
of wind or grunt of voice
separate parts of one articulation than two separate articulations.
Double explosives present some

tives are more easily managed.
Lat. pecco,
puppis,
diflficulty;
pronounce as in Eng. book-Case
double
frica-
(not as in peccant)
LATIN SYLLABLES
31
CHAPTER VI
LATIN SYLLABLES
64.
syllable has been defined as " such a
sound or
succession of sounds as can be uttered with a single breath
But this definition is not quite accurate, since it

does not cover the case of certain initial syllables. In the
impulse."
words scena, spatiiini, std.tim, scrlbd, spien, spretus,

stru5, the initial consonant combinations sc, sp,
spl, spr, str
scr,
st,
cannot be uttered with a single breath impulse
mute (c, p or t) is
made for the sibilant s.
the same slight degree,
in each case the articulation of the
preceded by a distinct breath
The
definition also
fails,
efifort
though
in
in the case of certain consonant combinations at the end of
syllable, e.g.
urbs
however,
this,
is
of little importance.
perfectly satisfactory positive definition of a syllable
to formulate, but the exact significance of the

term will be seen from what has already been stated and
from what follows.
is difficult
The number
65.
(1)
as
si,
it
of syllables in a word.
in Latin consists of as
many
syllables
contains "vowels separately pronounced."
Thus
rgo, cdr5n&, basilica, delibrar6, liberalitatd,
contain
(2)
vowels.
all
Any word
1,
2, 3,
For
4, 5,
this
and 6
purpose diphthongs count as single
Thus cauponae contains three
diphthongs are long,
them.)
syllables, respectively.
it is
syllables.
(As
never necessary to " mark
"
LATIN SYLLABLES
32
When
(3)
and u are used as consonants, they
cannot, of course, either alone or with other consonants,
Thus iocose (= jocose), and uanitas

(= vanitas) are words of three syllables maior (major)
and u616 (volo), of two syllables. So in such words as
suade5 (three syllables), lingua (two syllables), qui
form a
syllable.
(one syllable), where
and u
is
(Note that 1
On the rare
a consonant.
as consonants are not marked.
occasions
when the u
vowel,
is,
it
in such a
marked
of course,
suadeo is a
thus suadgo contains
word
as
four syllables.)
The
66.
(1)
vowel
words into
division of
consonant goes with the following
single
de-li-be-ra-re, &-ca-de-mi-a.
in a consonant, this final consonant
vowel
last
pal-li-dus,
is
If a word ends
attached to the
ci-te-ri-6r.
Two
(2)
syllables.
consonants coming together are divided
ma-gis-ter, hdr-ta-bi,n-tur,
dis-cus>si-
o-ne.
The
(3)
last rule
speech and verse)

sonants
z and
But
(4)
if
holds good, so far as sound (in
concerned, in the "double" con-
is
erulto
= ek-sul-to gaza = gadza.

;
the two consonants coming together form
a glideless sound-combination (a
r,
1,
or consonantal u, see
mute
18),
or f followed
by
both go with the
following vowel, the underlying principle
being that
such a combination cannot be divided.
The
possible combinations
pr, tr,
cr
46
br, dr,
gr
may
be
e.g., S-pri-ciis,
e.g.,
fifteen
classified thus:
pa-trls, v6-lii-crls
IS-t^-brS, quS-drii-p6s, 5-grI-c6-ia
LATIN StLLABLFS
7
pi, cl
910;
U 12
e.g.,
m^nl-pltis, sS-cla-stis
bl,
gl
e.g.,
pu-bll-cfis, rg-gln-tl-nS
ft-,
fl
e.g.,
va-M, rg-M-Qs
e.g.,
lln-guSS rg-qul-es-co^ de-sue-ttis'.
1315; gu, qu, BU

(5)
33
In poetry,
many
of these consonant combina-
tions are frequently divided, the object being to lengthen
the preceding syllable

r&, d,g-ri-c6-la.
This
e.g.,
is
pit-ris, vo-Iuc-ris, la-tgb-
a very important point, as will
presently be seen.
(6)
If the combinations are themselves preceded
by a consonant, the combination
still
begins the syllable,
the other consonant going to the preceding syllable
in-quam, lin-gua.
in&-gis-trd, De-cem-bris, ciil-tri,

(7)
Rule (1) holds good
for
compounds
d,-di-g5,
d-bae-ra-tiiB, though for the convenience of showing

clearly their origin, the
compounds are often written
etymologically (ad-igo, ob-aeratus).

(8)
If,
however, the
first
twelve consonant com-
binations in Rule (4) are brought about by composition,
the mute and the liquid must be kept separate, being

ob-la,-tiiB (not 6pronounced in separate syllables
bla-tUB) sub-la-tUB (not BU-bla-tUB) i,b-ii-pi-o (not
ab-rum-po (not a-brum-po). In such
d,-bri-pi-d)
cases there was very little fusion between the mute and
:
the liquid.
The
prefixes
formed distinct words
for
very long time.

(9)
The consonant combinations
be confused with those

1
w.
See 37, 38.
Rule
in
"
64 must not
in
(4) of this section.
See 43.
The
See 47.
34
LATIN SYLLABLES
former, unless initial, are invariably divided'; the latter
(See Rule
are never divided, except in poetry.
The general
(10)
rule in the case of
two consonants coming together

falls
before
the last consonant;
(5).)
more than
that the division
is
or,
if
the last two
consonants form one of the combinations in Rule

before
that
combination
planc-tus,
(4),
cdns-cls-co,
pirs-pi-ra-re, pa-Ius-tris, ins-tru-o^

This
'
is in
accordance with the weight of authority, despite Priscian's
rules.
From
the fifth century of our era onwards, the
Boman
granunarians
down the rule that all the consonant combinations which can begin
a word must be joined to the succeeding, and not to the preceding syllable
But the rule can hardly be correct. It
(e.g. a-pice, a-nmla, ca-atra).
seems probable that certain main principles of phonetics were imperfectly
laid
understood.
vowel
final short
is rarely
sq, 8t, gn,

tr,
but they are rare
lengthened before two consonants at the
Instances are occasionally found before sp,
beginning of the next word.

;
and instances are
still
more rare before
sc,
pr, br,
Ac.
The question
difficulty.
An
no means free from

on the subject appears in Classical
January, 1906. It is by Mr Walter Denniaon
of syllable division in Latin is by
interesting paper
Philology, Vol. i. No. 1,

and is entitled " Syllabification in Latin Inscriptions."
The evidence given
by the large number of Inscriptions examined, as regards (1) the division of

words between lines ; and (2) the separation of syllables by interpunetuation
(dots), goes to show that the consonantal groups were divided so that the
second syllable always began with a group that could begin a Latin word.
For
c-t,
(1)
the percentage
was about 80
for (2)
much
Thus we have
larger.
K-n, p-t, inp-a, &c.
There are two apparent anomalies

(i)
m-n and -mn are both found, though
:
(ii)
groups beginning with
m-n
is
more frequent
are divided after the
a-p, a-t, a-c,
though these groups may begin words in Latin.

Also na-t and na-tr (one instance) are divided as here shown.
LATIN SYLLABLES
35
syllable may begin with any vowel, diph(11)

thong, single consonant, or any one of the consonant
combinations in Rule (4). Initial syllables may begin

with all these, and, in addition, with those in 64 also
with gn, but only in the older language.
;
The final syllable of a word may end in a

(12)
short or long vowel, a diphthong, or one or more consonants men-sS,, men-sa, men-sae, re-git, re-gSnt.
:
An aspirated mute though spelled with two

must be regarded as a single consonant: d-le-
(13)
letters
ph&8.
In
fact,
the aspirate
may
for all purposes of syllabification'
generally be ignored
pul-ch6r, pul-chrS,
p&-rd-chus.
Length of
67.
The
syllables.
correct pronunciation of Latin depends very
upon an accurate knowledge of syllable length. The

length of the syllable and the length of the contained vowel
are two totally different things, and must not be confused.
There is an old rule to the effect that a vowel short by
nature may become long by position^ before two consonants.
largely
In face of this evidence we cannot say that a

is wrong.
syllable division like
la-strao
It is
probable that the syllable division in groaps containing
was not
Also the fact that in verse some poets lengthen a final

syllable that ends in a vowel, and some keep It short, shows that the
had a certain mobility, and it was this uncertainty which led most verseperfectly fixed.
writers to avoid the collocation of short final vowel +ao, ap, &a., altogether.
'
The iama poiitione
oi positu
meant
originally "
by convention."
32
LATIN SYLLABLES
36
But a vowel cannot become long by

short, it
remains short
If
position.
it is
naturally
neither the length nor the pronuncia-
In the word nicto the e is short

comes before two consonants does not
lengthen it. On the other hand the syllable nee- is long.
Some of the older grammars were in the habit of marking the vowels in such syllables long, e.g., necto, but this
tion of the vowel changes.
the mere fact that
it
was a serious blunder'.

A vowel may be long before two consonants, e.g., e
Its
in mensa, but that is because it is naturally long.
"position" has nothing whatever to do with its length.
The syllable men- would be long whether the vowel e was
long or not in the word mendum, for instance, the syllable
men- is long, though the vowel is short.
;
As
68.
Ellis
pointed out, the most
artificial
part of
Latin quantitative rhythm consisted in taking the short
vowel or short syllable as the unit of length and supposing
was always of the same length, and that the long

Nothing
of the kind is likely to have occurred in speech or declamation.
Still, to begin with, this artificiality must certainly
be aimed at, because we have nothing like it in English
that
it
vowel or syllable was exactly twice that length.
except in baiTed music.

to destroy our
rhythm
'
is
This
modern Western
is
necessary
if
only in order
which quantitative
habits, in
not known.
Even Kennedy
did this (Grammar, pp.
reasons of typography that
make
it
6, 513).
difficult to
No donbt
there are
mark the length
of the
than by patting a long mark over the vowel. In gradnses

it is the established practice and in these it ought not to mislead.
But the
practice is open to objection, for it certainly does mislead,
even students
syllables otherwise
of older growth.
LATIN SYLLABLES
37
69. Every long vowel must then, in practice, be regarded as equivalent to two " times (pendulum beats, for
instance), and a short vowel to one.
Musically, a long vowel
may be compared to a crotchet and a short vowel to a
''
quaver.
consonant,
when
must be regarded
considered,
as equal to a short vowels
Syllables are long or short according as they require " two

times " or " one time " to pronounce.
But in estimating the
length of a syllable, the length of time occupied by the

consonants preceding the vowel is disregarded, even when
there are two or three, as in sta-, Bcri-.
counts, or if the vowel be followed by
and
The vowel alone

a consonant, the vowel
that consonant.
70.
a long
In the following words, a figure S is placed under

and a figure 1 under a short syllable
syllable,
a-nl-ma
ro-bo-ro
2222
de-sue-tu-do
The length
le-ga-ti
splSn-dl-de
222
212
11
as-p6c-to
222
de-flft-gro
212
212
212
dis-cl-pli-nft
of the final syllables of a word should be
considered in connection with the initial syllable of the

following word
homlnis anim&
= hd-ml-nl-s&-nl-m
homlnis corpfts
= ho-ml-nls-c6r-ptts
11111
112
1112
2
judlcftt
Snlm sensiis = ju-dl-cS-tS-nlm-sen-stis

2
This mle is suEiciently exact for its purpose, thongh it ignorea lefineOf course the length of a consonant differs in different positions,
ments.
e.g.,
in bub&Ung and bui&ous, clear
and
ta2c.
But
see the next paragraph.
38
LATIN SYLLABLES
b5nft statera'
= bS-nas-tft-te-ra
bona creta
= bo-na-cre-ta
b5no pflSr5
= b6-n6-pii-S-r6
The
12
112
(cf.
(cf.
64)
Rule 4)
66,
12 112
final syllable of a final
considered short
12
word
in a clause
may be
ends in a short vowel, or a short vowel
if it
and consonant'. If it contains a long vowel, it is, of course,

long in any case. (With reference to "elision" of vowels,
see the next chapter.)
71.
Briefly:
(a)
(6)
A syllable is short when

A syllable is naturally
it
ends in a short vowel.
long
when
it
contains a
long voweL
(c)
syllable is long hy position
consonant.
But
when
it
ends in a
as a syllable cannot end in a consonant
unless the next syllable begins with a consonant, we

have the practical rule that a syllable containing a
vowel immediately followed by two consonants
is
by position (the glideless combinations in
Rule
4,
ehort vowel followed
by
66,
long
of course excepted').
'
This
is
initial $t or
a very rare type of sequence (i.e.

any other combination of 61).
paragraphs of footnote
But the length
2,
a,
Cf.
the second and last
pp. 34, 35.
was unsettled,
and a short syllable might be used as a long one. At the end of a sentence,
a syllable was no doubt often lengthened. Whenever there is a pause of
necessity, a short syllable with that pause may count as a long one.
So at
the end of a verse.
8 But these combinations often make "position" in poetry.
(See 66,
Eule 6.)
'
of a final syllable having a short vowel
LATIN SYLLABLES
The explanation
lies
in
39
of the lengthening
"by
position"
the phonetic division of the syllables.
consonantal combination shared between two syllables
lengthens
the preceding syllable

a consonantal
combination not shared between them leaves the preceding syllable short. Strictly, it is not the consonants
;
which add length, but the pause which separates them.

In the following line
DividimuB muros
note that the
first
-inuB
et
moenia pandimuB urbis
is
a long
So with the
-inu(B) short.
5;yllable,
and the second
syllable -ter in the following
line
Tractus uter plures lepores, uter educet apros.
72.
The
artificial
nature of the rule laid
down
in 69,
of taking the short vowel as the unit of length, of
i.e.
regarding the long vowel as equal to two of these units,

and a following consonant as equal to one, is seen in such
a word as potenB, in which the relative length of the
4 (po-tens). This wpuld be absurd.
syllables might be 1
however constituted, must be
all
syllables,
In practice,
regarded as having one of two time values. They must be
either short or Imig, and the short syllable is always approxi:
mately half the length of the long

time, a succession of syllables
cession of crotchets
syllable.
is closely
As regards
analogous to a suc-
and quavers, the crotchets and quavers
Lealways accurately preserving their relative lengths

ga-tl or tri-gin-ta takes twice as long to pronounce as
1
It
nuty be stated that there ie evidence of variable or intermediate

But any attempt to imitate them is bound to end in
length of syllables.
failure.
ELISION OF VOWELS.
40
One
Tk-g&-r& or l&-cri-in&.
make
SLURRING
commonest
of the
the accented syllable of a word long, and
syllables short,
is
A long
syllable
which
is
" long
to
the other
The
without regard to their real length.
question of accent will be considered in Chap.
Note.
faults
all
9.
by position
"
scanned long in Latin poetry, for, even though it

have a short vowel, the consonant group following reIf
quires a certain period of time for pronunciation'.
such a syllable contained a naturally long vowel there
is
would be really extra length, but there
Roman
taken in
the second syllables of
67
69
In the English
"Await
and
say,
Cf.
Lindsay, 142,
YII
SLURRING
ELISION OF VOWELS.
"
of,
supra.
CHAPTEE
73.
no account
calSsco and modestus, both
being treated as long syllables.

with
is
poetry of the different length,
lines,
alike the inevitable hour,"
Hurled headlong flaming from the ethereal sky,"
the vowel sound of the word the does not entirely disappear.
It is
touched upon very lightly.
So
a
final
The Romans much
in Latin.
vowel with the
correct enunciation,
initial
when a
disliked the clashing of
vowel of the next word.

final
For
vowel stands before an
should be lightly and shortly pronounced,

and run on to the following word. The final vowel should
not be elided entirely, though this is much preferable to
initial vowel, it
'
Cf. the first
paragraph on
p. 39.
ELISION OF VOWELS.
SLURRING
41
having hiatus, a practice which, as Professor Postgate points

most explicit language.
out, Cicero protested against in the
Thus
is
pronounced
ill'ibit
quaeque ipse
quaequHpse (not quaequ'ipse)
Roma
RomP'urbs (not Rom'urbs)
ego eo
ille ibit
Although
urbs
eg''eo
(not eg'eo)
in these cases the final vowel
form a
not, of course,
not three.
syllable.
(not
is
ill'ibit)
just heard,
Roni'ûrbs = two
does
it
syllables,
(Cf. 52.)
74. The
slurring takes place even
the second word
bona haec
if
= bon'^haec
the letter h begins
anima
horret
anim'^horret.
When
entirely
cumat
the two vowels are the same, the
omnis
ergo
ergomnis
first is
suppressed
er-gom-nis ;
Marcella
= Marcellamat = Mar-cel-la-mat.
The
efifect is
exactly
that of a single word.
So when a
vowel
final
is
followed by a diphthong be-
ginning with the same vowel: contra audentior = contraudentior
con-trau-den-ti-or.
75.
In such a line as the following

VenSrftt insano C&ssS,ndrae Incens&s fimorS
(Ve-nS-rS tin-sa no CSs sSn-dr'in cen-sti-sa mo-rg)

|
e,
the slurring of the ae in
Cassandrae
naturally causes the
almost or entirely, to disappear, and there is a tendency

a to fuse with the 1 and form a diphthong. Thus,
for the
although
the vowel
it is
if
the
first
long or
if
vowel which must always be elided,

a diphthong,
may tend
to
(so to speak) a following short vowel, especially

is
in an unaccented syllable.
if
overwhelm
this
vowel
In the above case the vowel
42
ELISION OF VOWELS.
SLURRING
must be distinctly pronounced with its own (short) quantity,

the a of the ae in Cassandrae Ueing only just heard.
(The i, though
by position.)
But
an
in verse
short,
final
initial short vowel.
a long vowel
is,
more
more
Much
is
much
some are
e and I elide
than & and 6 before a short vowel.
as hiatus
was disliked
it
was permitted
found occasionally even in the same verse,

Posthabi|ta coltt|iss6 SS,|mo.
Here the long o of
76.
followed,
than a
less easily sacrificed
in the case of short vowels,
But
Samo
is
Hie
end
and it
at the
of one verse before an initial vowel in the next verse

is
long
is not often followed by

what might be expected, as
than others; for instance,
easily elided'
freely
in a syllable
long vowel
This
naturally,
Even
short vowel.
happens to occur
e.g.,
lUliis
ftrmS-.
preserved.
in prose the general
rule
must always be
the final vowel, whether long or short, of the
viz.,
word must be only lightly pronounced and run into

vowel of the following word. This must be done
such a way that the two syllables are reduced to one.
first
the
in
No
initial
must be taken, even when a long vowel or a

Only the quality of the elided vowel
must be heard. The resulting quantity must be entirely
extra time
diphthong
is
elided.
that of the initial vowel of the following word.
The same
rule should be observed in verse.
Exceptions
are numerous, but these are always pointed out
by good
commentators.
1
The word "elision" must not be taken

" Slurring "
in its literal sense of absolute
perhaps the better term.

very nearly disappears, bat not quite.
suppression.
is
The "
elided " vowel
QUANTITY
77.
is
43
Within words, open concurrence (hiatus) of vowels
very common, but even here
it
is
avoided in
dififerent
ways:
(1)
The former vowel may be

nuUus.
(2)
The
often disappears with
adverto = animadverto.
(3)
Two
cut off: ne-ullU8
its
vowel
animum-
concurring vowels unite into one long vowel
pr6-61es
= proles prd-emo = promo.

;
So much was hiatus disliked that, in poetry, without

written contractions, two vowels were sometimes uttered as
forming one syllable, as aurea, omnia.
CHAPTER
VIII
QUANTITY
78.
Accurate pronunciation of Latin
is
out of the
question unless the natural length of every vowel is known.

The difference between long and short vowels was un-
doubtedly very marked in classical Latin, certainly as much

as between such English words as rod and rode or shin and
machine. It has already been stated that there is evidence
of some syllables being of intermediate length, and it is
quite possible that when a long vowel was followed by two
more consonants the syllable was longer than a long

But on these points the evidence which we
can regard as unquestioned is very scanty, and it is admitted
or
vowel alone.
by competent scholars that the only practical course
is
to
QUANTITY
44
consider the short syllable as the equivalent in length of a

short vowel, and the long syllable that of the long vowel, the
latter
being twice the length of the former.
It
is,
however,
that very few syllables in Latin consist of a long

Most have a consonant prevowel, or of a short vowel, only.
fact,
ceding the vowel,
Long vowels
e.g.,
no-ta.
are indicated thus:
Many
vowels thus
marked,
at most, only
or,
a, e, &c.
a, e, &c.;
and short
books leave short vowels un-
mark those
be thought
likely to
long.
Beginners in Latin would do well to mark the long

vowels of every
much
new word.
Such a rule would save them
trouble later on, and would go far to ensure an accurate
pronunciation.
many grammars
It
is
to be regretted that the writers of
are so careless in this matter of
marking
the vowels, and few even of the best dictionaries consistently
mark the vowels
79.
in final syllables.
The principles underlying composition, substitution,
and other changes, all influencing vowel length,

form too vast a subject to be dealt with in detail here. But
numerous facts may be grouped together in such a way as
to make it possible to formulate useful rules as aids to
contraction,
memory. The following rules deal with the length of vowels

in
I.
Final syllables
(a)
ending in a vowel
(6)
ending
in a consonant.
II.
Syllables not final.
Monosyllables are included under " final " syllables.

will
(As
be seen, the vowels in most monosyllables are long, but
there are numerous exceptions.)
QUANTITY
I (a).
45
Final syllables ending in a vowel.
A final is short:
Except (1) Abl.
mensa, audacia, maid,

sing, of
nouns
(2) Imper. sing. act. of

(3)
quid, &c., &c.
ita,
mensa, pdeta, &c.

verbs
ama, da, &c.
Most indeclinable words in a
erga, quadrd-
ginta, &c.
final is short : lege (from
amdrt,
Except (1) Abl.
sing, of
(2) Imper. sing.
(3)
lex),
KgS (from
lego), timstS,
-quS, -n?.
nouns fade,
Adverbs from
dXe^.
act. of
verbs
adjs. of 1st
marie, tSne, &c.
and 2nd declens.
recte, alte, docte", &c.
(4) Monosyllables
me,
1 final is long
Except' (1)
(enclitics
excepted):
e,
de,
se, te, ne.
plebi, doll, ilU, consUli, di, mi, si, &c.
nlst, qu&sl.
So hSdie but Mri.

So fire, firme, pgrSgre, ohe, vaMe, vere, bnt b^nS, mSK, infUrrU, supgrne.
> The short t and o forms in many of the exceptions given, and likewise
the short e in bgng and tbAK, represented originally long Towels or diphthongs,
The change was due to a process
e.g. mSM from madd, blni from bine.
known as iambic shortening. In words of two syllables the first of which
1
'
was
short, there
was a marked tendency to shorten the final syllable if long,

This was not a mere
to
rhythm from
.
that is to change the word
matter of poetic usage but a characteristic of common speech. In isolated

forms, such as those mentioned the tendency had full sway, and the short
vowel is prevalent from the earliest period. In mihi, &o., the form with the
short vowel became established but the poets continued to use also the old
form miki, &o., at
all periods.
As Professor Postgate points
out, the shortening of -6
verbs, spread to cretics, as deslno {-o in Tibullus),
analogy, to verbs of different scansions.
and
began in iambic
these, possibly
by
QUANTITY
46
(2) mXhi, tibi, siH, Ubi^, ibi
mihi,
final is long
(commoner than
&c.).
multo, juvo, amo, bono, carbo, origo, do,
no, &c.
Except" (1)
ego, cttd, dild,
Sometimes
(2)
cMo, modo, quomodd.
scio, nescio, piltS, volo,
and other
verbs when used parenthetically, and often

in poetry.
Sometimes nemo, homo.
(3)
Naso, and other proper nouns often

end in o.
(4) Scipio,
In post-Augustan writers, virgo, ergo, odd,
(5)
porro, immo, sero, intrS, &c.
U final is long
Y final is short
Summary
tu, dictft, diu, &c.
Tiph'p, ckelp.
final
and y are
I
i,
o,
and u are long" ;
final a, e,
short.
Final syllables ending in a consonant.
(b).
(1)
The vowel
except
in a final syllable ending in
any consonant
s is short''.
Examples
-b
-c
-d
'
db, 8b, sub.
(But
fdc, nee, donSc.
lac, d%c, sic, hoc,
due, hue.)'
ad, id, quid, quod, illud.
Note oompoonds iblnam,

See note 3 on p. 46.
:
itilvis,
^
but HblquS.
This will saggest a simple mnemonio.
' Exceptions are fairly nnmerous, thoagh most

of the important ones are
mentioned, except those in -x which are many.
' Also iUic and ittic.
Note that most of the examples of long yowels
before c are shortened words, not original monosyllables.
QUANTITY
-1
47
fil, mSl, vel, nihil (or nihil), Hannib&l.
(But
sal,
sol, nil.)
-m: diem, hast&m, unitm, quid&m,
Srni'Sm,
mSnSdm,
&c., &;c.
-n
&n, in, agmtn, &c., &c.
(But
en, quin, sin, non,
lien, siren.)
-r:
fSr,
Ur,
pSr,
calcar,
cdr,
(But Ldr, par,
Hector, &c.
amahUHr,
ver, cur,
r^gitor.
fur, crater,
asr, aether.)
-t
dat,
St,
tot,
flU,
&mat, dnrnvit, &c.
contracted perfects,
fax, vix, pix, m5x, nox, nUx, dux,
fiUx, &c.
(But pax,
(But note
audit, &c.)
it,
rex, vox, lUx,
criix,
vindSx,
Umax, radix,
cervix, &c.)'
(2)
The vowel
-as
is
vds
in a final syllable
long
ending in
s varies as follows
dmas, terras, mdnSds, misSras, as, fas,

(But ands, vds (v&dis), lampadds,
(vasis), &c.
Ili&s, Grdterds (ace. pi-)-)
-es
is
long: sedes, morns, audires, videres, es {Mo),
pdries, &c.
Except
(a)
Nouns
in Ss (with consonant stems)
having genitives in^

(a) -Stis, e.g., seges, tSgSs, inteipres, &c.
(/3)
-itis,
e.g.,
eques,
pMSs, milSs,
c5mSs, hospes, &c.

(7) -idis,
e.g., obses,
(b) Ss {esse), pSnSs,

^
Note that x
a compound sound. The vowel usuallj follows

pdx=pdc-s rex=ree-s (reg-is) &c.
quadrupes, dMes, aries, paries, and many others come
is really
the quantity of the stem
Kote that pes,

under the rule and not under the exceptions.
'
praeses.
Arcades, crdteres{v.pl.).
QUANTITY
48
-is is short
ensis, regis, ciicumis, cUciimeris,
utUis,
dicSris, &c., &c.
Except (a) Dat. and Abl.
plu. mensls, voMs, quis,
bonis, &c., &c.

(b) Ace. pi.
So gratis, /oris.
omnis, finis, &c.
(c) Indie, pr.
2 sing, of I verbs, as audis;
also possis, velis, nolis, &c.
(d) Subj. perf. 2 sing., as
amaveris (but
sometimes amdvSrisy.
Samnis, Quiris, Eleusis, Scdamis, &c.
(e)
-OS
is
long
flos, 6s (oris), sacerdos, ventos,
Except compos, impos, 6s

chlam'gdds (Gk. gen.
-UB
is
short
opits, intiis,
mos, &e., &c.
Argos, Delos,
(ossis), ex8s,
s.).
amdmus, amabdmus, gSwOs,
hostibus, &e., &c.
Except
(a)
Gen. sing, and
Nom.
Ace.
pi.
of
nouns, as exerdtus.
(6)
Nouns
in -us
which increase in the
Gen. with a long penult

palUs,
virtus,
incus,
tellus,
sSnectus,
juventus, jus, &c.

(c)
Some Greek nouns: Sapphiis (gen. s.),

Panthus
II.
Syllables not final.
(n. sing.),
There
is
far
&c.
too great
variation in the quantity of vowels in syllables not final for
any approach to completeness to be possible, and

a thoroughly good dictionary becomes absolutely necessary.
rules with
'
The converse
is
the case with the indio. fat. perf.
QUANTITY
few helpful rules
49
may however
be given
(1)
All diphthongs are long'
(2)
Vowels which have originated from contraction
caudae, foediis, &c.
or represent diphthongs are long
(3)
co-go from
c5-&go;
momentum
ntMl'
exlstimo for ex-aestimo, &c.
The quantity
for
nil for
of the radical syllable of a word
generally preserved
position,
m6vi-mntum;
in
even when the vowel
mater, rndternHs
comchanged:
is
scribo, scriptor
cldo; c&do, incido;
is
derivation and
caedo, in-
But exceptions
&c.
are
numerous.
(4)
vowel before another vowel
The
ratio, &c.
intervenes
Exceptions
is
short
piUs, vld,
rule holds good even if an h
contrdhU, dShlsco.
(a) Genitives of pronouns, &c., in
-iiis
Ullus, alius, solivs, totlus, umv^,&c.
But the
tives is
in several of these geni-
sometimes
short.
(6)
Old genitives of the
(c)
Genitives and datives of the
first
declension,
like aulal.
when a vowel
precedes
fifth,
dlei,
but
fidH.
(d)
The a
or e before
in -ius:
cf.
'
The diphthong ae
in proper
Odius, PompeiHs.
58.)'
Also
in prae in composition becomes short before a vowel,

Practically the e is suppressed,
noonced pra.
' Or nihil.
vocatives
(But
plebeiiis.
as in praeesse, praeire.
nouns
The
oai and FompSI
and prae
is
pro-
are of similar scansion.
HIDDEN QUANTITY
50
(e)
The
the syllable fi of fid, ftunt,
i in
But
fvebam, &c.
-er
(/) In
(g)
dl'iis (adj.),
The vowel
short before
oer, aera, Aeneas.
i is generally
consonantal
pMiui
{derivo),
(But
KuiUm
triuiiim (trivium).)
of a vowel followed
a case of special difficulty, and
long before
deriuo
(petivi), &c.
(bivtitin),
The quantity
it is
fltrem, &c.
by two consonants
is
dealt with in the next
is
chapter.
CHAPTER IX
HIDDEN QUANTITY
80.
In the vast majority of English words a vowel
by two consonants
followed
I5st,
not
lost
when we remember
should expect
An
is short.
additional effort
is
We
This
concert, not concert.
is
say, for instance,
we
exactly what
" the law of least effort."
required to effect the pronunciation
of the two consonants, and the shortening of the vowel
sound
is
the natural consequence.
exceptions.
We
But there are many
do not say, for instance, told, but told
not
bind but bind.
So
it
was
in Latin.
In the great majority of Latin words,
a vowel followed by two consonants
is short.
Still,
there
is
a large number of exceptions, probably more than we are
acquainted with,
The quantity
of
for
our knowledge of them
many
is
incomplete.
vowels followed by two consonants
therefore said to be " hidden,"
though the term
"
hidden
is
"
HIDDEN QUANTITY
is
51
sometimes applied, to the quantity of those long vowels

by two consonants) that are known to us.
(followed
Clearly, then, a vowel before
naturally short,
two consonants
is
sometimes
sometimes naturally long, although the

vowel is, as a rule, long. (Cf.
syllable containing such a
Chap.
It
6.)
short, as is
is
not correct to pronounce the vowel invariably
commonly done.
be long, as it
pronounced long.
to
is in
Whenever the vowel is known

number of cases, it must be
a large
In some words the length of the vowel before two conThe phonetic law of least
effort leads us to think that in most of such cases the vowel
sonants has not been determined.
was probably
all
short.
The
safe plan, therefore, is to
pronounce
such vowels short unless they are known to be long.
81.
The quantity of a long vowel

is known with certainty in a
consonants
followed by two
large
number
of
In other instances there is a strong probability

of the vowel being long. In others, authorities are divided
In still others, there is merely a slight prein opinion.
sumption in favour of a long vowel. The following rules
are in accordance with the weight of authority.
instances.
Vowels are always long before na, nf: dens,

mens, frons, pons, mons, mensa, ingens, sapiens, amdns,
monens, nolens, conspicio, constituo, consulo, consto, expdnsum,
(1)
mdnsum, pansum, pro/nsum, scdnsum, censum, defensum,

extensum, mensum, pensum, sensum, pmso, pinsitum, sponsvmi, tonsum,
tunsum,
conficio, confiteor,
&c., &c.
conflrmo, con/undo, confluens, infdns,
Infellx, Infensus, Inferior, Infimus, Infra, &c., &c.
It has already
been said
( 36,
40) that
before s or f
42
HIDDEN QUANTITY
52
was probably a mere nasal, lengthening the preceding vowel.

Hence some scholars, in pronouncing these words, prefer to
omit the n altogether they would pronounce consul either
as co-sul or as co-sulK
There is no objection to this, provided the vowel is kept long.
When a g becomes c before t or b the preceding
(2)
vowel becomes long actum (from ftgo), lectits (ISgo), rectum
(rggo), tectum (tSgo), afflictum {a.^\go), frlctum (frigo), suctum
;
(sugo)
= recsi
so rean
(rSgo), texi
tecsl (tfigo)
also rex,
lex, &c.
Likewise fractum,, pactum, tactum (from frango, pango,

tango)
are a
but fvctum, pictum, strictum
Contrast factum, (from

(paciscor),
(fingo, pingo, stringo)
doubtful.
little
fftcio),
tr&ctus (trfiho),
nactus (nanciscor), pactu^s
enSctum (enSco), sectum
(sSco),
inspSctum (insplcio), vSctus (vSho), dictum (dico), victum

(vlnco)", ictu^ (ico), frictum, (frico),
(duco), in which the c does not
Note
cdctum (c5quo), diictum
come from a
g.
be noticed that the lengthening

of the vowel before ct or cs, when the c comes from
a g is a change in perfect harmony with the ordinary
:
It should
laws of phonetics.
(3)
the vowel
(cingo)
As might be expected from

is
usually long before

(fingo)
fvnay\,
strvnxl (stringo)
the preceding rule,
net and nx
exstinctum,
cinctum, clnxl
exstvnxl
tlnctum, tlnxl (tinguo)
(exstinguo)
functus (fungor)
junctum, junxi (jungo); punctum (pungo); unctum, unxi
There
(unguo).
(distinguo).
IB
some
doubt about pldnctum,

and distlnctum, dlstlnxl
On the other hand, there is some authority
planxl (plango)
'
is
slight
pinxl (pingo)
a nasaliBed
o.
gut victum (from
vivo).
HIDDEN QUANTITY
for ndnctus (nanciscor)
53
sanctum, sdnxi (sancio)
vinctum,
vinxl (vincio), although in these cases the c does not come
from a
g.
When
the phonetic laws of the Latin language are
better known,
it
is
highly probable that the present
doubt about flctum, pwtum, strictum, planctwm,

pldnxl, pinxi, distmctum, dlstliuci, wjll disappear also
slight
that the
long vowel
now suggested, by one
authorities, in nanctus, sanctum, sdnxi,

will
or two
vmctum, vlnxl
prove to be short.
A vowel is long before ps and pt when these

(4)
come from bs and bt scrlpsl, scrlptus (from scribo) nwpsl,
nuptum (nubo). In these cases note the long vowel in the
;
present indicative.
The vowel
(5)
is
long
if
in nominatives ending in -x, -ps, -bs
long in the
other
cases:
lex
Cyclops
{legis),
(Cyclopis), plebs (plebis), &c.
There was a general tendency to lengthen a vowel

before r followed by another consonant, and in the case of
many words the pronunciation of the vowel as a long vowel
became the recognised one. This is true of Mars, Marcus,
Ma/rdus, Lars, ordo, ordior, omo, forma, quaHus, and several
(6)
others.
The vowel is usually long before gn dgnus,

(7)
mdgnus, mdgnitudo, stdgnum, regnum, segnis, dignus, slgnum.
Ignis, lignum, pugna, pugno, agndtio, agnomen, dgnosco,
;
cogndtus,
cognitus,
cognomen, cognosco, igndvia,
IgnSbilis,
Ignominia, Ignosco, Ignotus, igndrus, prlvlgnus, &c.
In some of these cases it has been established beyond doubt that the vowel was long by origin, and it
HIDDEN QUANTITY
54
known that the pronunciation of the vowel as a long

vowel existed in other cases. But there is still much
is
doubt about Roman general practice.

A vowel is long before sc in verbs in -boo
(8)
pdsco, noLscor, eresco, compesco, seised, nanciscor, vlciscor,

obliviscor, proflclscor, nosco, dgnosco, cognosco, Ignosco, &c.,
There is a little doubt about disco, compesco and pdsco,

&c.
but they probably follow the general rule.
Shortened perfect forms in -&BBe, ease, -ISBe,
(9)
-aBti,
-esti,
&c.,
-iBti,
have a long vowel before the b:
amdsse, amdssem, delesse, delessem, audisse, audlssem, amcbstl,

amdstis, delesti, delestis, audistl, audistls, nosse, nosti, &c.
In a considerable number of words, the long a

(10)
which comes before two consonants is explained by derivaThus we have frdtriclda from frdter, mdtrimonium
tion.
So with the long o in such a word as
from mater.
solstitium from sol.
The long a
-dhrwm, &c.,
from a principal; thus we have
fldhrum from fidre doldbra from doldre labrum, lavdbrum,
and lavdcrum, from lavdre ardtrum from ardre simulacrum
(11)
is
in derivatives in -dtrum,
also obviously derived

;
from simuldre
&c.
(12) The quantity of a long vowel in a principal is

generally the same in compounds, derivatives, and parallel
formations. Thus we know that the a in dctum, is long,
and we may therefore infer dctio, dctor, addctum, co-dctum,
perdctvm, &c.
Similarly /rojidis' {frons) implies frondosus;

vendo implies vendidi and venditum ; nuntio implies nwntius.
; purgo, pHrgdtio ;
forma, formo, formula; flrmus, flrmo, flrmdmervtum
So with
scrlptus, scriptor
'
Lewis.
drdeo, drdesco
(But doubtful.)
HIDDEN QUANTITY
usurpo, usurpatio ;
narro, narrdtus,
55
narrator;
osculum,
oscular; oscito, osdtdns; prosperus, prospero; prospicio, pro-
spectus
rectum, directum, perrectum, surrectum ; and so on.
In such Greek-borrowed words
(13)
pegma, Metro-
as
dorus and metropolis, the long vowel before the two consonants
is
83.
simply the reproduction of the Greek long vowel.
Long vowels
pairs of consonants,
are frequently
and
ciple or rule concerning
make a
advised to
all
them
is difficult,
in
fact, to
before other
is
discoverable^
The reader is
them
himself as he comes across
authoritative
events no exhaustive
met with
occasionally happens that no prin-
list for
No
in his reading.
it
list,
list
of such words, or at
has yet been published.
prepare such a
list,
It
in view of the
The following
about certain words.
to give
will
suffice
means
exhaustive,
but
it
list is by no
may
of
forms
he
of
variety
some
idea
the
the reader
expect to meet with.
Long vowels before glideiess combinations of con(i)
difiference of opinion
sonants.
br.
(Cf.
66
(4).)
crabro, candelabrum, ebriu^, libra, cribrum, sobrius,

lubricus, delubrum.
(See also Rule 11.)
(See also Rule 11.)
cr.
mucro, involucrum.
dr.
dodrWns.
gT.
dlgredior (and therefore dlgressus, &c.), vegrandis,
and possibly frdgrdns^

tr.
atrium, cldtrl, Idtro (bark), Idtrlna, cicatrix, quln(See also Rules

qudtrus, nutrio, nutrlx, verdtrum.
10 and
13.)
the other hand, the origin of the long yowel is often obvious, e.g.
purgo from purigo, Idrdum from Idridum, Idmna from lamina, pulvilltu dim.
" Lewis.
(But very doubtful.)
of pulvinui, and so on.
1
On
HIDDEN QUANTITY
56
bl.
publicus (publico, publics,
cl.
periclitm:
gl.
digladior, iugldns.
(ii)
(See
Long vowels
&c.).
before certain s groups of consonants.
64.)
8C.
osculum
(osculor), oscito
(oscitans), vdsculwm, esca,
sescunda, sescentl, prlscus, roscidus, muscus, and
probably
sp.
(See also Rule
lentlscus, vlscus.
8.)
prosperus, prospicio, aspemor, sospita and sospes
probably
crisptis
possibly suspido (noun) and
suspicio (verb).
St.
fdstiis (court-day),
tillus, vastus,
(nefdsttis),
ndsturdum,
pdstum, pastor, pds-
bestia,
comestum,
festiis,
palimpsestus, sestertius, easistimo, mlstum (mixtum),
plstum, plstor, prlstinus,
tristis,
ostium, prostellum,
prosterno, prostibulum, crUstuin, crusta, frustv/m,

fustis, Justus, paluster, pustula, rusticiis, ustus.
Perhaps vestigium, vestibulum,

also
Btr.
Rule
dlsto, tostum.
(See
9.)
rdstrum, bimestris, semMstris,
jnstrilla,
plstrinum,
sistrum, rostrum, frustra, lustro, lustrum, inlustris.
Long vowels
(iii)
before various other consonant com-
biuations.
ct.
lector, rector,
Ivbctus,
also
plectrum, vlctum (vivo), fructus (fruor),
ructus, strHctum.
Perhaps fluctus.
(See
Rules 2 and 12.)
gm.
agmen, frdgmen, frdgmentum,, prdgmaticus.
11.
mdlle, vallum, catella (chain), anguilla, catillus, bovillus,fovilla, mille, pistrilla, pulvillus, stilla, suilhis,
villa, corolla, olla, nolle, ullus, nullus.
Probably
HIDDEN QUANTITY
also favilla, hillae, ovilltis.
vSllo, pSllo,
paMla,
57
(Contrast p&lleo,
vSlle,
tUillo, villus, pdUeo, pUllus, &c.)
Im.
pulmo.
Is.
solstitium (see Rule 10).
Probably also alsi, fulsl,

mulsl (mtilgeo), indulsl (note the g in the present
indicative of these four perfects, and cf Rule 2,
(Contrast falsus, pulsus,
note).
ir&lsus,
Twitlsi
mps. dempsi, compsl, pj-ompsi, sumpsi (note the
origin
(mtilceo), &c.)
It.
ultimus, ultrd^.
tab.
comburo'.
mn.
Idmna.
of the
long vowel, in the present indicatives).
(Contrast confimpsl.)
mpt. demptus, comptus, promptus, sumptu^.

conttmptus.)
There is doubt about
(Contrast
emptits (co-
emptus, redemptus, &c.).

princeps, quincunx, uncia, semUncia.
nc.
net.
unctio, &c.
nd.
prendo
(See Rule
(for pre-hendo),
3.)
vendo
(for
venum-do) (see
77 (2) and 79, II, 2), nondum, vmdedm,

quindecim, undemgintl, nUndinae, vindemia, and
possibly /rondis' (frons).

rally short
combination
ng.
before
:
(But the vowel
this very
common
nonne.
quinque, qulnquiSns, quitiqudgintd, &c.
'
&c.)
nongentl.
nq.
'
gene-
amdndus, monSndus, Undique,
nn.
But both very donbtfal.

and Heinichen, p. 602.
Lewis.
p. 760,
is
consonant
Heinichen.
(But very doubtful.)
See Lindsay,
'
p.
594, Stowasser,
See p. 54, footnote.
HIDDEN QUANTITY
58
emtio
nt.
nuntius (nuntid,
(for co-ventio),
before nt the vowel
is
&c.).
nearly always short
(But
amWnt^,
rexissSnt, &c.)
nx
quincunx, conjumc.
pa.
repsi, redpse.
pt.
scriptor, reptum,
also
Rule
(See also Rule
(See also Rule
3.)
4.)
sceptrum, nuptiae, allptes.
(See
4.)
rd.
orca, siirculus.
(See also Rule 6.)
ardeo (drdesco), bardus, Idrdum, ordo, ordior, prl-
rg.
jurgo, purgo {purgdtio)
re.
mordium.
(See also Rule

;
6.)
and possibly pergo, exper-
giscor.
rm. forma, firmits.
(Contrast yormido,
(See Rule 12.)
formica.)
m.
uma.
vemus, hornus, orno, and perhaps
(Contrast
superûs, &c.)
rp.
Hsurpo.
IT.
ndrro.
rs.
arsi,
(See Rule 12.)

(See Rule 12.)
drsurus,
horsum, intorsum, introrsum,
orsus,
proreum, qitorsum, sinistrorsus, rUrsum, swrsum.

Probably also spdrsi, spdrsum, mersi, mersum,
tersi,
tersum, conspersi, conspersum, Ursi.
also fors, forsit,
morsum.
(But
t6rsi,
Possibly
cursum.)
rt.
qudrtus,furtim,furtum. Possibly also /orte, /orfosse,
sq.
sesqui.
SB.
mdssa, fortdsse'.
(But fortitudo,
fortuna, fortuitus.
pdssum,
'
Roby has
p. 91)
p. 90).
'
cessi,
all
(S.
9.)
Probably also
cessum, sessum, consessum, gressus,
avUlnt [Grammar, Pt.
he also has regent
But in
(See also Rule
fortiter, &c.)
i.
p. 227)
and amant {School Orammar,

i. p. 26, and S. G.
G. p. 16) and regint (G.
verbal endings the vowel before nt is invariably short.
Stowasser, Heiniohen.
ACCENT
69
fissum, scissum, fossum (note the d in the pres.

indie, of these verbs) ; missum, esse (eat), jussi,
jilssus, ussi.
tt.
mitto.
X.
See Rules
5 and 12
rexi,
teoci,
dlxl.fixl, -fl%x%,
nMximus. Possibly also traxl, flexl, -frlxl, rixa.

(But alUan, nexi, p&cl, v&ci, c6xl, stritxi, Jl&mim.)
juxta, mlxtum.
xt.
(iv)
large
2, 3,
duxl, luxi, fluxi, suxi, fvmim, luxus, niams,
vlayi,
long vowel before two consonants
number
of proper names.
found in a
is
Here are a few
Puhlicola, Socrates, Africa, lolcus, AquilKus, Illyria,

Pollio,
Lemnos,
Temnos, Glytemnestra,
Vlpsdnius, Norha, Lycnrgus,

Gressa,
Thressa,
Manlius,
Ndmia, Mars,
Tartessus,
Telmessus,
Lars,
Gnossits,
Polymestor, Sestius, Sestos, Ostia, Praxiteles, Oxus,

Roxarie,
Amazon^.
CHAPTEE X
ACCENT
83.
economy,
When, in English, we pronounce such a word

we are said to " accent " the second syllable, or,
as
in
other words, to give that syllable a greater emphasis or

stress
^
than the other three.
But the
precise nature of this
Professor Sonnenschein objects to tbe recognition, in ordinary practice,
of hidden quantities, mainly on the groands of conflict of opinion, of in.

sufficient evidence,
and
of unnecessary additional difficulty.
the Class. Assoc., Jan. 1912, pp.
1913.
8792, and
Class. Sev.,
See
May
See also Oanon Sloman's reply in Class. Sev., Not.
Professor Buck's reply in Class. Rev., June 1913.
Proc. of
1912 and
1912,
May
and
ACCENT
60
it
additional " force " or " loudness " does
simply means
seem
To say that
never been satisfactorily analysed.
stress has
entirely to
meet the
There
case.
is
not
probably a further
element, that of musical pitch.
Everybody knows that the sound of a violin-string or of

is due to vibrations brought about by
bowing and striking, respectively. If we tighten the string
a pianoforte-wire
or the wire,
we
increase the rapidity of the vibrations, with
the consequence that the pitch of the sound

larly,
in
the pitch of the
human
voice
is
Simi-
is raised.
raised
by an
increase
the tension of the vocal chords.
When
same
two notes from two different sources are of the
pitch,
their rates of vibration are the same.
If a
tuning-fork, or a pianoforte-wire, or a violin-string, or an
organ-pipe, or a tongue of a concertina, all produce the
note as the
human
voice, it is
same
because the vibrations of the
tuning-fork, of the pianoforte-wire, of the string, of the air-
column, and of the tongue of the concertina,
all
vibrate at
the same rate as the vocal chords of the singer.
Changes of pitch
by leaps or by glides.
in the
human
voice
may
proceed either
In singing, the voice generally dwells
without change of pitch on each note, and leaps upwards or
downwards
to the next note as quickly as possible.
differences of pitch are easily measured.
voice only occasionally dwells on one note and
moving from one note
is
constantly
to the other, so that the different
notes are simply points between which the voice

tinually gliding.
difficult to
An
An
The
In speech, the
is
con-
These differences of pitch are exceedingly
measure accurately.
absolutely level tone hardly ever occurs in speech.
approach to the level tone
may
be heard in the word
ACCENT
well as an expression of
intonation
may
statements, as,
musing
61
or meditation.
" rising "
be heard in questions or doubtful hesitating

are you ready ? A " falling " intonation may
be heard in answers, commands, or dogmatic assertions, as

in Yes, I am.
A "compound rise" may be heard in such
a sentence as take care ! when uttered wamingly a " com;
pound fall " in oh ! oh really ! when implying sarcasm.

But in ordinary speech, the extent to which (if at
all)
pitch enters with stress into the so-called accent in accented

syllables of English polysyllables,
It
is
uncertainty which
we
feel in
it is
very diflScult to say.
which creates much of the
this particular difficulty
estimating the value of the
available evidence in connection with accent and accentua-
Our knowledge
tion in Latin.
of Latin accent
is
slight,
authoritative evidence being meagre, but this has not pre-
vented many well-known investigators from adopting a very

dogmatic attitude on the subject. Many of their mistakes
are due to misleading associations with the term accent as
commonly used
in connection with English.

" accent " one of pitch or of stress ?
84.
Of the musical
no doubt at
quite clear'.
all.
Was the Latin
character of Greek accent, there
is
Dionysius of Halicarnassus makes this
From him we leam
speaking
is
in songs
and on instruments
a musical one too
that "the art of public
for it differs
from that used
in quantity, not in quality.
For in public speaking, words have also melody, rhythm,

modulation and propriety. In speaking then also, the ear
is delighted with melody, is impelled by the rhythm, and
especially longs for propriety.
'
In a treatise
irepl
The
difference
jvydâeuis dro/iiriav.
is
merely one
ACCENT
62
The melody
of degree.
of speech
is
measured by a single
is as nearly as possible that called

proceeds to identify Greek accent with
musical interval which

a Fifth."
pitch,
And he
and does
so in explicit terms'.
As
Professor Postgate
points out", the limited influence of stress on Greek phonology
may
well be due to the dominance of the pitch accent

which subsequently attracted it and was then absorbed
into
it.
In
fact,
there was, in course of time, a profound
change from pitch to
stress,
and the musical charactar of
accent seems to have disappeared altogether.
In Latin the problem
is
more
difficult.
A fifth
century
grammarian remarks' that the accented syllable in a Latin

word is the syllable which could be heard at a distance
when the others were inaudible. This, however, tells us
little, though perhaps it suggests stress-accent rather than
pitch-accent. Yet there can be no doubt about the existence
of a pitch-accent, and Brugmann, amongst others, explicitly
recognises that musical accent was still alive in classical times.
In Roman speech, as in Greek speech, there appears to have
been a rise and fall of pitch, though this, of course, was a
gradual slide, and not a leap through an interval as in
singing*.
And there is the clear and direct evidence of
authoritative Roman writers to the effect that, in what
they regarded as the accented syllable of words, there was
by Professor Postgate, see " Flaws
Academy, 1907-8, pp.
1845. Cf. Ellis, Quan. Pron. of Latin, pp. 2729. (Dionysius also
clearly stated that there was a strict observance of quantity in prose.)
" Classical Beview, 1899, Notice on Bragmann.
'
For
Ellis's translation, as corrected
in Classical Eesearch," Proceeding! of the British
'
*
Lindsay, Latin Language, pp. 17, 152.

This is proved incidentally by Vitruvius, Archit.
v.
4,
Aristoxenus (Postgate, Class. Rev. Vol. xa. p. 364 footnote).
translating
ACCENT
a distinct rise of pitch.
63
This evidence
it is foolish
we may
to controvert, in spite of difficulties
explaining other phenomena of the Latin language.

instance,
for
the question
solidum and soldum.
syncopated
of
to try
experience in
Take,
doublets',
these that Latin accent was entirely a stress-accent,

Professor Postgate clearly shows', these
e.g.,
from
It is useless to try to prove
for,
as
doublets will be
produced whether the accent is pitch or stress, and if rapid

speaking shortens a word the qualitative predominance of
the musical accent is just as effective a protection to its
proper syllable as
stress-accent.
is
the quantitative predominance of the
True these doublets do
about the syllables that

syllable of
weakened
sprang into being.
lie
went
force
tell
us something
outside the main accent.

to the wall
The
and a doublet
Hence, when we see solidum syncopated
and calidus to caldus, we can infer that the

was pronounced with less force than the respective
to soldum,
syllable
li
final syllables
dum and
dus.
The educated pronunciation
of Latin, at least so far as
quantity was concerned, was powerfully influenced by the
Greek, and no doubt in both languages there was a tendency
way to the stressBut there is nothing
for the pitch-accent gradually to give
accent and ultimately to be lost in
it.
show that, at least in classical times, differences of pitch

and differences of stress were not absolutely independent.
They may not have been, but our positive knowledge of the
whole subject is of the slightest, and the delimitation of
pitch-accent and stress-accent is a matter which investigators
to
have
still
to determine.
See Postgate, Class. Phil. Vol. in. No.
'
lb.
a.
LindBay, op.
cit. p.
in Herviathena for 1908, pp. 202
150.
foil.
1,
pp. 99100,
and note 2
Cf. also Professor C.
ante.
Ezon's views
ACCENT
64
It
is
true that ancient writers on Latin
grammar were
seldom Romans by birth and that they were usually Greeks,

and no doubt, as Mr Lindsay says, the latter would have
the same diflSculty in describing the Latin accent as a
Frenchman
in describing the strong accent in English.
But
the explicit testimony of the ancient writers, considered

as a whole,
must stand
first;
and
modem
apparent inconsistencies of language, second.
inferences from
That the two

is no doubt,
things are at present frequently in conflict there
but when a reconciliation between them takes place, there is

certainly a balance of probability that the former will prove
to have been right and the latter wrong.
Would those who
argue to the contrary accept from Macaulay's New Zealander
an English essay of a twentieth century school boy, or a

cockney coster song, which he had unearthed in forgotten
London, as evidence to controvert Sweet's statements about
the present-day spelling and pronunciation of English
Not the
least of our difficulties is to decide precisely
what the ancient
writers
" pitch " in connection
meant when they used the term
Of the nature
own spoken language
ill-informed as we are
with accent in speech.
of this pitch as they used
it
in their
they were, presumably, at least as

of the nature of pitch in our own.
Present-day phonetic
experts are by no means agreed as to the relative pitch of
when we pronounce such a word as

One well-known professor thinks that pitch and
the difierent syllables

economy.
stress are absolutely
independent and that in every word
the syllable of highest pitch
is
the
first,
while other com-
petent authorities maintain that the stressed syllable of a

word is always the one of highest (or lowest) pitch. The
point
has not yet been experimentally determined satis-
-factorily,
and hence
rival
hypotheses prevail.
Once we
ACCENT
65
begin to dwell, as in singing, on a particular syllable of a

spoken word, the pitch is changed almost unconsciously.
Correct measurement
is
On
when,
thus very
may
mately the phonograph
difficult,
though
ulti-
help us to solve the problem'.
the whole question, Professor Postgate thinks that

ever, the facts
if
which bear upon the case are duly
ascertained and co-ordinated, the probability
is
come
that the out-
be something like the following: The separate

syllables of independent words in Latin had uniformity
neither of pitch nor of stress. The syllable which received
the greatest force might be the one which had the highest
will
pitch or
it
might
From
not.
the variation arose, especially
in the earlier period, fluctuations
and anomalies of quantity,
as, for example, syllables which were neither short nor long
in the strict sense, and syllables with different quantity in
different forms of the
same word.
Under Greek
influences,
these variations were reduced until there was an approxi-
The
mation to a condition of nearly uniform stress.
preponderant stress (the " main stress-accent) in the case
of polysyllables tended in the time of which we know
''
anything to move toward the end of the word, though it

never passed beyond the second word from the end, and
thus to coincide with the main pitch-accent. It was not till
after this movement was completed that the joint-accent lost
musical character and survived as a stress-accent alone".
its
'
Di E. B. Edwards has
difficult subject,
that
when
the voice
already
made
using the phonograph.
It
certain investigationa into this
seems to be an established fact

much less than when
the voice is tired, differences of pitch are
is
fresh
and
vigorous.
On
the remarkable differences of pitch in
Chinese speech, consult any educated Chinaman on speech "tones."
See
also Bibliography, Vol. 43.

'
Accent in Latin, Close. Phil. Vol. in. No.
w.
1. Professor Postgate
gives
ACCENT
66
85.
that
is
There
is
general agreement about one point, and
very marked stress given to
that the
accented
and German was quite unknown in

Latin.
On the other hand, there was something of a more
definite nature than is found in French, where the even
gliding over the dififerent syllables of words sometimes
makes it difficult to distinguish any accent whatever except
syllables in English
at the extreme end of a sentence.
In practice the accepted rules are
Accented syllables in Latin must be pronounced

much more gently than accented syllables in English.
(i)
Unaccented
(ii)
syllables in
rapidly slurred over as

Particular attention
of every syllable.
much
just as
is
must be paid
An
must not be
Latin
so often the case in English.

to the proper length
unaccented long syllable takes
time to pronounce as an accented long
and twice as much time as an accented short

There must, in fact, be a much more even
distribution of force over the syllables of a Latin word
than in English, and the accent must never be allowed
syllable,
syllable.
to destroy the proper length of the unaccented syllables.

(iii)
No
must be made to pitch

The usual (and
of pitch supposed (by some
conscious attempt
the accented syllable in a higher key.

unconscious) difference
authorities) to exist in ordinary English speech will

suffice.
me
the following useful references on Oreek and Latin accent
of HalicarnaBsns, de Comp. 11
Aristoxenus)
25 and 31)
xii. 10. 33.
Vitruvius, de Archit.
Cicero, Orator, 68;
For other
111), especially vols. 22, 26, 28.
v.
Quintilian, Imt. Or.
i.
Dionysius
2 (based on
5. 22 aqq. (esp.
4.
references, see Bibliography (pp. 109
ACCENT
It
67
must not be thought
that, because there
was
possibly a fixed difference of pitch between accented
and unaccented
syllables, there
was in Roman speech

That Virgil in
anything of the nature of chanting.
make
his hearers conscious
of other musical qualities than
mere rhythm there
reading the Aeneid would
can be no doubt, and
it is
possible that, in poetry,
chanting of some form usually

place of reading
but
it
if
not always took the
requires considerable imagina-
tion to think of Cicero intoning

forensic orations.
his political
There was probably at
modulation of voice in the case of
least as
Roman
and
much
orators as
in the case of the best speakers of the present day.
In what follows no distinction is made between

and the " circumflex " accent. If due attention
be paid to the proper length of syllables and of vowels,
86.
the " acute
''
especially of vowels in final syllables,
it
is
distinguish between the two kinds of accent.
unnecessary to
At
all
events
the only other distinction, that of rising and falling pitch,

is
probably outside the range of the possible practice of any
but a few experts.

PoBitioii of the
Accent.
General Rules:
I.
(1) Words of two syllables are accented on the
first
(2)
dilrd,
minsd, bdnis, mddd, fdrte.
Words of more than two syllables are accented
(a)
the last syllable but one (the penult) if

that is long' legdtl, ndtiir&, amdbas, amdre,
On
It
has already been pointed out
67
et seq.)
that a long syllable often
contains a short Towel.
52
ACCENT
68
ingintes, m&gisf&r, &cpdndunt, compirtii/m, Idti-
hortabuntUr, amahdmiis, debilitdrS.
titdo,
(b)
On the
but two (the antepenult)

short ndtlo, cdmprimd, pittas,
last syllable
if the
penult
is
qudmSdd,
cdptent,
Idcriniai,
rlgtret,
vSlilcHs,
tin^rae, mdUlssimae, offindMre, drndvero, Todnu-
potuMram, h5rtdr6rm,n%,
Srds,
nihildtninios.
Subsidiary Rules:
II.
(1)
Compounds
are accented according to the general
no matter whether their parts can be used
rules,
words or not: ddfero, cdnflcU, tSrri-
as separate
gena,
res
ddmodum,
dnhil'tLs,
pdstmodd, respiiblicd (or
pdhUcd).
But
in non-prepositional
compounds
oi facio,
such as c&lSf&cio, t^pefdcw, Mnefdcio, mansuifacUd

(which were originally written in separate parts),
the accent
is
always on the verb; thus c&l^fdcU
(not c&Ufdcit)
so in
the passive,
calS/tt,
&c.
But prepositional compounds follow the general

rule:
(2)
dfficit, cdnffcit, &c.
Sometimes a
or there
accent
is
become
(a)
a long penult
In such cases the
final syllable following
is lost,
is
contraction.
retained on the syllable which has
now
final.
Adjectives in -as (-dfis) denoting one's native

place.
The
original form
was -dtta which was

Maecends
contracted to -dts and then to -as
Arpinds,
prlmds.
Ldrmds
so
also
nostrds,
vestrds,
ACCENT
(b)
69
Shortened perfects of the
fourth
audit (for audtvU), muntt,

(c)
Shortened imperatives
conjugation:
pgttt, dbit.
&ddtlc (for Mduce), pro-
die, benedic.
(d)
Shortened demonstratives
(e)
Similarly tantdn
(for
illic (for
twntdne),
audtn {audisne), vidin
satin
connubia servas'' (Virg. Aen.

(3)
Genitives and Vocatives in
preserve their accent as
endings in ii
a short penult
The
{s&tlsng),
3.
So
"
pottn
Pprrhin
319).
of nouns in -ius
they retained their

and ie; thus the accent appears on
:
if
Virglll (for Virgilll and Virgilil),
dvidl, MerciiH, V&Ur%.

(4)
istic.
(for vidin, vidisnS),
{potisnS), dntehdc (dntShdc^), &c.
The
UUc^,
So tngSm
(for ingSn>Ci).
enclitics -que, -ne, -ve, -ce displace the accent
and attract it to the final syllable of the words

to which they are joined^
Thus hdminilm, but
homin'&mquS; Ul% but illicS; dAHs but dMsvS;
cdunctH but c&nctdnS?
So with -Twit, -pte, and -dum: thus, Sgo but
SgdmU ;
ndstrd but nSstrdptS; rispici but r&pi-
cidHm.
And
with the separable words inde and quando
thus, Sartnde (not SxindM); 8cqu&ndo (not Scqudndo).

(5)
Such words
as libet, licSt
and quXdlm are sometimes
called enclitics, since, like enclitics proper, they
may be
cases,
1
rule.
Bat in
ThUB
attached to other words. But in such

owing to their own quantity, the accent is
earl; Latin such forma were accented according to the general
bona, bonaqnl ; Itmind, liminaqui.
ACCENT
70
same whether they be considered

proper or as parts of a compound:
the
scilicet,
as enclitics
quibuslibSt,
So with
or qudnddquid^m.
qudnddquidem
quantumvis.
(6)
In certain words the particle -que has become inseparable; it is no longer a conjunction and is
Hence the accent
therefore no longer an enclitic.
follows the general rule
denique, itaqut, utique,
undiqu^.
But
in
some
of these cases the -que
occasionally do duty as an enclitic.
itdquS
= and
so
and
as
'iltiquS=
(7)
itaque
may
still
Hence
therefore.
uti,qvM= in any case.
Throughout the declension of utirquS, the accent

Thus we have not only
falls on the penult.
atSrque and utrumquS (according to rule), but
also utrdque.
(Cf.
Rule
So with plerusquS,
4.)
plerdquX, plerumquS.
(8)
Prepositions are seldom accented or stressed' in any
way, even in English

ately preceding their
altogether.
'
For
and in Latin when immedinouns they lose their accent
all practical
purposes a preposition
may be
regarded as compounded with
noun.
Thus pSr
We may
hostes is not
pSr
its
hdstes
following
but pSr
be said to "accent" a syllable of a word, bnt the additional
force given to a
word as a whole
is
usually
known
as "stress" (in French,
Hence, although the term "accent," as ordinarily

hardly applicable to monosyllables, many monosyllables naturally
"oratorical accent").
used, is
receive stress or oratorical accent, according to the subject
But words like prepositions and conjunctions are

words to receive stress in ordinary speech.
and the speaker.

than any other
less likely
ACCENT
71
sUpra mdntSm is not supra mdntSm but

supra mdntem. They are not, however, real compounds, e.g., mt&r nos is not intSrrws but intSr n6s.
In every case the preposition loses its accent
hdstes;
altogether.
however, a preposition follows
If,
own
retains its
accent
ti
its noun it
qudcum, quce-
sinS,
prdpter.
But ciim
after a personal
pronoun
enclitic
is
thus nobiscum, voMscttm.

(9)
Conjunctions beginning a phrase do not take an

accent; but
not beginning a phrase they are
if
accented according to general rule
quid vilUnt; but

(10)
The
relative is
tii
sM
Odlll, si
qudquS, v6s igitur.
unaccented, the interrogative ac-
quo die rSdiit (on which day he returned)

qud die? (on which day ?).
cented
(11)
The accent
in the case of
mentioned in
pronunciation
is
when, as
is
many words
of the type
The normal
66 (4) (5) varies.
Id-t^-hrd, md-ni-plus, &c.
But
often the case in poetry, the consonant
combinations are divided in order to lengthen the

preceding syllable, the pronunciation
ma-nip-liis^.
^
time.
i.e.
the
mute
is
In such words as
is
la-Uh-rd,
dgricdld
and
pronoanced with the preceding vowel, adding a nnit of
Contrast the division of the word latebras in the two following lines
ant |te-re-|bra-
re ca-lva-
tnm{le-vis|haut|alI
su-te-iri et| tempi ta-
|tra |la-te-|tiraa|
jam
|re la-|teb-{ra>.|.ien. 2.
|quae-|ri'ti-
In the same verse Horace has nig-ris, ni-groque; Ovid,

Virgil, pa-tris, pat-rem; Lucretius, pa-tribus, pat-res.
|ma-
go.
38
|ên.l0.663
volu-cri, voluc-rii;
ACCENT
72
p&tris, the division of the consonant combination
although changing the length of the syllable

thus &-gricannot, of course, aflfect the accent*
;
and ag-ri-c6-la ; pd-tris and pdt-ris. (It

has already been pointed out that in these cases
the length and the pronunciation of the vowel
c5-ld
remains unaltered.
It
is the syllable,
not the vowel,
that is lengthened.)
of the pronominal genitives in
The pronunciation
(12)
-ius
may
also vary.
Thus we may have
(See
79, il,
Exception
a.)
U-li-its or U-li-ils ; to-tt-us or
td-tX-'&s.
words of more
than three syllables frequently had a second,
though lighter accent; and even a third in the
It is probable that, as in English,
(13)
thus applicdrS, mllgni-
case of very long words:

t'Ado,
The
(14)
nohilltdMs, SxercUdtiSnibHs.
first
part of a
compound
especially
may have
retained to some extent the accent which

as a single
word
however, the English practice
If,
it
had
nihilSminHs, ammddvSrto.
the shortening of long syllables
at
is
all
closely imitated,
is inevitable,
and
it
has to
be remembered that, in Latin, correct syllable length
is
of
fundamental importance.
'
Since in the case of agricola, the penult
division
is
short with either
the antepennlt therefore takes the accent
the accent mnst in either case
fall
on the
method of
in the case of patrit,
first syllable.
(But this does not,
of course, apply to such a word as tenebras, the normal pronunciation of
which
Quint.
ia
1.
tinebras, but,
5. 28.)
when a long penult
is
required, tenSb-ras.
See
GENERAL REMARKS ON QUANTITY AND ACCENT
73
CHAPTER XI
87.
common in the "restored

new ") pronunciation as in the " English "
" old ") pronunciation of Latin.
They are principally,
The shortening of long vowels in all unaccented
(1)
Roman "
(the
Certain faults are as
(the
"
syllables.
(2)
The shortening
of all long syllables
which are
not accented.
(3)
The lengthening
of
short vowels in accented
syllables.
These
faults are such as naturally result from our English
habits of speaking.
sarily, to
The heavy English accent
tends, neces-
lengthen the accented syllable of a word and to
shorten the other syllables.
88.
It is a very
common thing
for
of,
a word as amavSrUnt, provided that the
c is
and
is
classical teacher
for example, such
to accept as correct the pronunciation
pronounced long
accented, the difference in the length of the
a and
the a and the proper length of the last syllable of the word
receiving no attention whatever. It is not pointed out that
the last three syllables of the word are of equal length and
that they take equal periods of time to pronounce. Thus
as a boy is introduced to verse
"
face
number
of
new rules " concerning quantity,
has
to
a
he
him
to
be largely artificial whereas
which
appear
to
rules
it
comes about that as soon
if
he had been taught to read prose in the proper way, most
74
of these " new rules " would be entirely unnecessary, and he

would discover the musical rhythm of the verse almost
unconsciously.
The word socQtas

mentioned in
87.
illustrates the last of the three faults
A beginner is
rightly expected to accent
but the chances are about even that if he

does not bring out a long clear English i^ he will be reprithe syllable
manded
for
ci,
not paying proper attention to his quantities
It
a rare thing to hear such a word pronounced correctly,

with the fourth syllable given twice the time of any one of
is
the three others, and with the vowels in the short syllables
given their correct phonetic length.
89. Typical words of two, three, and four syllables will

now be taken and their pronunciation illustrated by means
of English words.
The English words are carefully selected
in order to represent, as far as possible, the relative length
of the syllables in the Latin words.
short Latin syllable
example, be illustrated by an English word contain-
will, for
i or e or y, simply because such a

most suggestive of shortness. In a similar way, an
English word containing a long a, o, or u, or a diphthong,
sometimes followed by one or more consonants, will be made
ing a short vowel, usually
word
is
to represent the Latin long syllable.
Illustrative words of
kind are a great help in attuning the ear to accuracy of
quantity in a succession of syllables.
this
90.
Words
of
two
syllables.
It is obvious that as
and " short " syllables, we may have four

types of Latin words of two syllables, viz., a word of two
long syllables, one long syllable followed by one short, one
there are
i.e.
"
long
"
the diphthoDg heard in such English words as society, finally.
GENERAL BBMARKS ON QUANTITY AND ACCENT
75
short followed by one long, and two short.

These may be
conveniently named after the so-called " feet " in Latin, viz.,
The
The
The
The
But
Spondee (
)
Trochee (- )
Iambus ( -)
Pyrrhic (
as in cogi.
v/)
cwr&.
gSnu.
^^
bgnt
misleading to use the vowel marks (- and ^) to

represent syllable length, seeing that a long syllable often
it is
The marks will therefore only be

used to represent vowel-length.
Syllable lengths will be
represented by musical notes, a long syllable by a crotchet
contains a short vowel.
and a short
(J)
91.
dens.
(1)
The
syllable
by a quaver
The Spondee
M)
^Y.
c6gi, frae, h6rtoB, pru-
relative syllable lengths
may
be obtained by
imitating, quantitatively, the pronunciation of the English
words boar-hound.
These words
than, say, the words
tall girl,
illustrate the case better
for the
word boar takes a
slight accent, whereas in the second case the slight accent
Yet the word hound takes as much

The tendency with
this type is to shorten the second syllable, a tendency which
must be carefully guarded against.
falls
on the word
girl.
time to pronounce as the word boar.
> The crotchet and qaaver sae intended to lepieaent time periods only.
may, however, be mentioned that some anthorities write the accented
Thus
syllable of a word a perfect fifth above the unaccented syllable.
It
^
Â-rt
S
a
There
Greek.
is
ma
ve
rmit
not sufficient justification for this in Latin, though there
is
in
76
cur4, casta, reg, miU6. For

(J J*)
relative time, compare such English words as houses or fairy.
This is a very easy t)rpe. Care must, however, be taken to
prevent the final short vowel from becoming the indefinite
vowel sound referred to in 9.
(2)
The Trochee
(3)
The Iambus
I)
boniB, mei,
4m5, ginu.
For
compare such English words as sinews or volume.

more diflBcult than this. The first syllable though
accented must be kept short, and the second must be made
long. No fault is more common in Latin pronunciation than
to convert an iambus into a trochee: thus bonis is often
incorrectly pronounced bdnis suis, siiis mo, dmo and
so on.
Indeed, such words as suis and ml are difficult to
relative time
No
type
is
pronounce even approximately accurately.

(4)
The Pyrrhic (j*
h)
bine, pi&, c&pr&, propi. For
compare such English words as river or money.

Here again it is necessary to guard against lengthening the
accented syllable. The above words are frequently pronounced incorrectly, as bene, pi&, c&pr&, prdp. So with
relative time
all
Pyrrhics, so-called.
The somewhat common
practice of allowing boys to say
and conjugations with the accent on the

of the word leads to faults that are probably
their declensions
last syllable
seldom quite eradicated afterwards.
The accent
is,
of course,
never on the final syllable in the ordinary cases of declensions
and conjugations.
Contrast also
{J* J*)
h6stis
with
4m5
b6ms ( J*J)
(JJ*)
^ J)
(
;
with h6BteB
with
mensfi.
(JJ);
dm&t
(J J*)
&c.
^^); biniis
with
mensa ( JJ);
Words of three
92.
syllables.
77
It is obvious that as
each of the four types of dissyllables just mentioned
may be
preceded by another syllable either long or short, there
are,
as regards syllable length, eight types of words of three

syllables.
The
following table
shows (a) their metrical
names (which are used here merely

rence)
and
(b) their
(d) English
the time periods.

(1) Molossus.
time periods
words to
for
purposes of refe-
(c) illustrative
Latin words
illustrate as closely as possible
78
No. 4
is
easy
so is no.
7.
In no. 5 care must be taken to
So with
long.
The
of no.
8.
first
make the
last syllable
no. 6.
syllable of no. 7 should
Note that
it is
be compared with that
twice as long.
good practice to pronounce the suggested English

and then to pronounce the Latin words in as
nearly as possible the same time, and with a similar rhythm.
It is
words
first,
93.
Words of
types of trisyllables
either long or short.
of four syllables,
four syllables. Any one of the eight

may be preceded by another syllable,
There are thus sixteen kinds of words
all differently
length.
The
nounced
in conjunction
(1)
constituted as regards syllable
following examples^ should be carefully pro-
J J J J
with the suggested English words.
oratores,
defendissent,
"
our
slow
boar-
hound."
(2)
J* J J
ricflsivl,
ciiciirrirunt, " the
slow
boar-
hound."
(3)
(4)
(5)
" new cigar-box."

Siaboro,
concitari,
J ^J J
hj J v6ii6r4ri, monuissent, " the cigar-box."
noUtotfi,
commutare, "our white horses.''
J J J ^^
I*"
(6)
coronirt, amavistiB,
J*J J ^^
(7)
J J*J J*
(8)
J* J^J
(9)
J J J^J
litigitdr,
^ animare,
"the white horses."
d,pplicire,
"
voluissem,
" inelastic."
hereditas, mdllissimae,
"
no remainder."
mean
vulpicide."
1 The names of the corresponding metrical feet are not

given they are
hardly neoesaary and probably very few people ever trouble to learn them.
If required they can be found in any good Latin grammar.
;
GENERAL HEMARKS ON QUANTITY AND ACCENT
79
(10) J^ J J* J
Bev^ritas, ftmiverant,
"
the vulpicide."
(11)
radlitae,
perpltud,
"
white pinafore/'
(12)
J^^J^J
cel^ritas, monuferts,
"
the pinafore."
(13)
J J J*J
mlribile, interprttis,
"fine juniper."
(14)
J* J J* J*
J^J**J
pficlinia,
J J*J*J*mS.tiTi&,
(15)
(16) 'j*J*J*S*
amdbferis,
"
the juniper."
imprfinr,
"
white dimity."
m6m6ria, quadruped^,
Most of these require a great deal of

attention should be given to
It
is
all
' celerity."
practice.
unaccented long
Particular
syllables.
useful to compare no. 11 of this section with no. 1
The words noUtS and fQtilitaB are

pfonounced in exactly the same time.
The two intermediate syllables of the latter word take just as long to
pronounce as the middle syllable of the former. This should
of the last section.
be practised with the metronome or with the pendulum.

Other similar exercises may, with great advantage, be devised; e.g., memoiid, is pronounced in exactly the same
time as dico.
94.
If words
of two,
three,
and four
more than four
syllables
are
need
few examples are appended, to-
correctly pronounced, words of
syllables
A.
cause little difficulty.
gether with English words suggestive of relative length.
(1)
80
(5)
/J
J J-
J-
t&b^miculum,
" the
c611d.bdr4r,
(6)
"
(7)
(8)
(9)
green juniper."
" eighty-eight
IJ
95.
As
" stain
\J*JJ*
"
young
horses."
innumerabilitas,
^ ^
^ ^
***
J
horses.''
machinamSntdrum,
J/JJJ/
eighty young
inconBider&tissimiriim^
Gray bought
already stated,
-i-ii
on a white pinaiore.
it is
forty old
Roman
vases."
good practice to pronounce
in immediate succession pairs of words which are closely akin
Compare, for instance,

the word izploratorgs and the word innumerabilit&s.
as regards complete " time-periods."
PROSE AND VERSE.
Now
ACCENT
AND
81
ICTUS
pronounce, in time with the metronome, the English
words suggested
94 (1) and
in
(8),
and note how the accents

There ought now
naturally on the word boar and pin-^.
fall
to be no difficulty in pronouncing the Latin words so that
amount of time and so that

But the accent must be
The one important thing is time. Our
each syllable receives
its
the accent
proper place.
a very
falls in its
light accent.
English "sledge
must be
called)
"
rushing
" all
correct
hammer"
accent (as
it
has been so aptly
avoided, and our slovenly English habit of
unaccented syllables must also be avoided.
CHAPTEE
XII
PROSE AND VERSE.ACCENT AND ICTUS

96.
The previous chapters have
pronunciation of single words, and we
dealt mainly with the
now come
to the diffe-
rences between prose and verse.
These differences, in Latin

as in English, are characteristic, though they are by no means
But before it is
of the same type in the two languages.
possible to deal adequately with the question of the proall, by verse, it
English verse.
structure
of
upon
the
necessary to touch
nunciation
97.
in
is
of Latin, as affected, if at
The
its
syllables.
is
chief respect in which verse differs from prose
regular
succession
of accented
This regularity of accent
is
and unaccented
called rhythm.
passages are, however, often rhythmical
Prose
'
See footnote on previous page.

The nsnal convention has been adopted- of indicating the accented
syllable of a word by placing the accent mark over the contained vowel,
or, in the case of a diphthongal combination, over the second vowel of the
'
'
combination.
w.
PROSE AND VERSE.
82
ACCENT
AND ICTUS
The Dog-star and Aldebardn, pointing to the restwere h^lf-way up the So6thern sky, and
between them hung Orion, which gorgeous constellation
n6ver burnt more vividly than ndw
The barren and
"
less Pleiades,
glodmy Square
of Pegaslis
was creeping round to the
aw^y throdgh the plantation, Vega

sparkled like a lamp suspended amid the leMess
tre^s."
Far from the Madding Crowd.
"They mdved so gently that their fodtsteps md.de
n6rth-west
far
am6ng
the grotip and

For sh^ was dedd.
Th6re upon her little bed she lay at rest. The s61emn
stillness was no marvel n6w.
Where w6re the traces
of her edrly cdres, her sufferings and fatigues?
AH
g6ne.
But peace and perfect happiness were b(5rn,
imaged in her trdnquil beduty and profound rep6se."
no
noise,
but th^re were sobs
soiinds of gri^f and
mourning.
Th,e
Now
Old Curiosity Shop^.
with these passages compare the following
"Brought fr6m the wodds, the honeysdckle twines

around the porch, and se^ms in thdt trim pldce a plant
This passage from Dickens, though useful here for tllustrating the
is rather a shocking specimen of prose rhythm.
It was a
common fault of Dickens to write in this fashion he constantly fell into
1
point in question,
'metre.'
Even
of metre, but
in the first quoted passage, there is
Mr Hardy,
like all
good
more than a suspicion
writers, reserves snoh
exceptional occasions, and to the ear the result
is
usage for
not only unoffending but
is wholly pleasing.
Rhythm, whether in prose or in verse, is essentially a
recurrence of some kind of unit of length, but whereas in verse the unit is
a metrical foot and therefore easily defined, in prose it is more arbitrary,
follows no rule,
and cannot be defined
prose the succession of units

regularity.
Bible, e.g.,
is
Many admirable examples

Psalm
xovi., Ecol.
of Professor Brewster's
iii.
in specific terms,
sufficiently
of prose
18, and
uniform
rhythm may be found
especially 1 Cor.
The Writing of English.
though even in
to be suggestive of
xiii,
in the
See chap, ix
PROSE AND VERSE.
ACCENT
AND ICTUS
83
no Idnger wild the cultured idse there bldssoms, strdng

and will be so(5n roof-high; the wild pink
cr6wns the gd,rden w^ll, and with the fldwers are intermingled stdnes spdrry and bright, rough scatterings of
;
in health,
the
hills."
The Excursion.
This passage, although rhythmical like the
from them in that
it
may be
We
off in lines of equal length.
first
two, differs
and naturally marked
easily
thus have the following
arrangement, constituting "verse":

" Brought fr6m the wo6ds,
Around the
p6rch,
the h6neysuckle twfnes
and se^ms
pldnt no 16nger wfld
in thdt trim pl4ce
the criltured r6se
There bl6ssoms, string in he41th, and wfll be ao6n

Roof-high
the wfld pink cr6wns the garden wdli,
(
And
with the fldwers are Intermingled stdnes

Spdrry and bright, rough sodtterings 6f the hlUs."
|
In reading either prose or

sary;
verse, a frequent
in order to breathe, the utterance
pause
is
is
neces-
checked.
verse the pause takes place at regular intervals.
In
The length
of a line of poetry does not necessarily determine the length

of a sentence, but it usually does determine the length of a
phrase, for there
is
of a complete line.
some other
a natural tendency to pause at the end

If the line be long, a reader also pauses at
place, usually in or near the middle of the line.
In
the above lines from Wordsworth, the bars indicate the pauses.
The rhythmical arrangement of words measured

lines of regular length is known as metre.
98.
off in
When
reading verse the succession of accented syllables
seems to mark it off in equal steps if marching, our feet

would keep time with them. Hence, as many syllables as can
be grouped about a clearly accented syllable is termed a, foot.
;
62
PROSE AND VERSE.
84
The
ACCENT
foot is the unit of metre.
AND ICTUS
It usually consists of a
group of either two or three syllables, one of which is

A verse is a cycle of feet forming a line of poetry.
accented.
The number of feet in a line determines its metre. We
thus have the terms tetrameter, pentameter, hexameter, &c.
The following are examples of tetrameters
:
me
n6t in m6urnful numbers

unhdn- oured ind unsiing
Over the
roddways and
6n through the
1.
Trochaic:
Tdll
2.
Iambic:
Unwept
3.
Dactylic:
vfllages
I
4.
A mphihrackic
5.
Anapaestic^:
The
was a picture
fl^sh
for painters
to study
|
O'er the Idnd
the brdve
99.
of the free
and the h6me
of
In verse, the term
Sometimes
accent
is
ambiguously used.
refers to the stressed syllable in the standard
it
pronunciation of a word (in this sense

vious chapters), and
it is used in the presometimes to the marked pulse-beat
which the ear easily detects in the same place of each

metrical foot in a verse.
For the latter, however, the term
ictus is often employed, and will be so employed in what
follows here. In other words, ictus will denote metrical stress,
while accent will denote the stress given to a particular
syllable of a word according to the standard pronunciation.
In the above lines from Wordsworth the marks indicate

falls on the second syllable of all
the thirty-five iambic feet. In all the feet save one, ictus
the ictus, which, of course,
is
identical with accent.
the last line

syllable,
in the
The exception
is
the
word sparry, the accent
though the metrical
stress (the
first foot
is
ictus) is
second.
'
These terms
will
on the
be understood from 90, 91.
of
first
on the
PROSE AND VERSE.
A great
lOO.
AND ICTUS
modern poetry
deal of
of a varying character.
ACCENT
This
is
85
written in feet
is
a great relief to the reader
who soon tires of symmetrical verses of the " half-up, halfdown " type. Here is an example (the bars indicate the feet)
" Th^re be
none of
Beadty's
daughters
With a m^lgic
And
like
miisic
like thte,
|
6n the
w&ters
{
sweet vofce
Is th^
m^"
to
>
|
Variety
by giving the
also frequently obtained
is
unaccented
Thus, translated into quantitative terms, the
rubato.
of the following quotations gives us - ^
wv/|
two feet taken together

"Statesniffn
^-
,
1
and the
first
sixth,
Similarly, the total ictus in the
normal".
is
with sle^p-|les8 witch
(3)
"And wfld xash and

"He 16ft the up- land
(4)
"That
(5)
" A.qt6s&
(6)
"To
(2)
the total length of the two feet taken together
|,
being in each case normal.
(1)
ictus to
speak more correctly, by accento
syllables, or, to
and
steiid-|fa8t aim."
Ingram.
f-|vy s^r-|pentlne."
Idwns
and
Shdley.
Myers.
^-|rene air."
thou,
dead c6r3e,
again
darker
Shakespeare.
|
Longfellow.
than dedth or night."
its (i-|tique p6r-|tic(5."

I
/6r-|give wr6ngs
in c6OT-|plete ste^l."
Shelley.
ictus are shown in

upon
a word which, from
italics.
not
ought
to be stressed
certainly
the sense of the passage,
The unaccented
syllables taking the
Sometimes the ictus
"Calm dnd still hght on y6n great plain." Tennyson.

"Land 6f\ the moiin-|tain dnd the flo6d." Scott.
"Grows thi salt 14 v- lender that Idcks perfdme." Crahhe.
1.
2.
3.
1
falls
Metricians differ in opinion about the pointing of this last line.

This distribution and compensation of ictus may be spread over a
whole
line,
and allows of almost
infinite variation.
PROSE AND VERSE.
86
101.
ACCENT
AND ICTUS
In the great majority of metrical feet in English

and accent fall together. But where they do
poetry, ictus
not
fall
ictus,
is
together, a good reader will never sacrifice accent to

if
he did he would nullify the poet's intention, which
to avoid tiring the ear
Nevertheless,
good
by a too monotonous sing-song.
reader always seems
to
make
his
audience conscious of both ictus and accent, the ictus being

only just heard, however, and clearly subordinated to the
accent.
is
The natural accentual
characteristic of every
word
thus preserved, the swing of the metrical feet being at the
same time gently
indicated.
Both are
on
easily detected
the stage.
We may now
it is
return to the consideration of Latin.
desirable to point out that there
is
But
only a very faint
between the construction of English verse and the

The preceding remarks on
construction of Latin verse.
English verse are in the main intended to prepare the reader
similarity
for essential points of difference, not points of similarity,

between English and Latin verse construction. It is quite
true that, even in English, we cannot have really good verse
without a considerable amount of dififerentiation between

word-accent and verse-ictus, though of course the diflference
is
a matter of art, or perhaps of artifice.
In Latin, however,
the distinction between word-accent and verse-ictus
is
of
the most fundamental character.
102.
words that
In Latin, as in English, the poet so arranged his

they produced rhythm, and in the
for the reader
main he used the pronunciation of daily speech.

was determined by relative lengths of
syllables,
But metre
and in the
construction of verse everything waa subordinated to quantity.
Whereas
in English a
rhythmic
effect
is,
as a rule, brought
PROSE AND VERSE.ACCENT AND ICTUS
87
about by reading a poem according to the accents in standard

pronunciation, in Latin the rhythmic effect is brought about
by a regular arrangement of long and short syllables. In
English we subordinate all questions of quantity to accent.
In Latin we subordinate both ictus and even accent to syllable
length, though neither ictus nor accent is lost.
It has already been stated that the exact nature of Latin
accent
we
uncertain.
is
use
In fact the exact nature of accent as
in English
is uncertain, though it is at least

and some authorities think it probable, that in an
English word the accented syllable is pitched higher (or
lower) than unaccented syllables'.
If, as some authorities
think, the Romans always chanted their verse, and always
chanted the accented syllable of a word on a higher note
than the unaccented syllables, we can form a pretty clear
it
possible,
idea as to the manner in which they brought out both ictus
and accent,
for in Latin, unlike English, ictus
very frequently
103.
fell
on different
Reference to
and accent
syllables.
91 (1) and 92 (7) will show that
the spondee and the dactyl have equivalent time-periods,
and
and
it is
good practice to read aloud a succession of spondees
dactyls, correctly
timed by the metronome.
We may
take a few lines from Virgil:

Musa, mihi causas memora, quo numine laeso,
Quidve dolens, regina deiim tot volvere casus
Insignem pietate virum, tot adire labores,
Impulerit. Tantaene animis caelestibus irae?^
Chap. X.
Cf.
'
The {nndamental mistake made almost
of Latin hexameters
is,
as
Mr Mackail
aniTeTsallj in the reading
happily puts
it,
to read
them as
though they were in triple (musical) time, whereas they are, of course,
in common time. Oace this fact is realised, the rest is comparatively easy.
PROSE AND VERSE.
88
ACCENT
AND ICTUS
These may be written to shew the feet, the quantities, and

(The six feet of each verse
Cf. 52 and 73.
the accents.
are
marked
off
by the longer
bars.
The
short bars indicate
half feet, in order that the time-relations of the spondees
and dactyls may be clearly

Mu-
|sS mf-
caii-
lii
sas
seen.)
nfl-
ml-nS
la^- so
vS-rS
cS-
Quid-|vS d6-
l6ns
r6-
v8l-
pI-S-
dl-
sua
I
In-
|sig-
n6m|
Im-
Ipti-lS-
rit
JTSn-
r6 15- bo- res

1
l^s-
tl-bfiI
rae
31I
PROSE AND VERSE.
ACCENT
AND ICTUS
89
not accent-rbythm, but quantity-rhythm, as determined by

syllable length, that is the important thing.
104.
In reading the lines quoted in the preceding
them first in monotone, with the

metronome, both accent and ictus being entirely disregarded
the primary consideration is time.
The second reading
section, it is best to read
should also be in monotone, but in this case

syllables
merely
might be pitched
for
accented
all
a somewhat higher key,
in
purposes of differentiation.
The
ear ought
now
to be able to detect the natural " swing " of the verse, and
When
the third reading can take place in the ordinary way.
make
this is done, it is probable that the ictus will
even
we
felt in
every
foot,
habits
make
this almost inevitable.
if
try to suppress
it
itself
our English
Both accent and
ictus
will thus probably be heard, the accent, however, being the
more prominent of the

wanted. (Cf 101.)
two,
and
this is exactly
what
is
105. As in English, sense-stress may sometimes fall

upon a monosyllable that does not take the ictus, and as in
English the ictus may sometimes fall upon a monosyllable
that ought not to be stressed at
all.
(Cf.
100
also 86,
II (8), footnote.)
106.
(1)
Each
The two important

syllable
accented or not
things to bear in
must be given
(2) the
its
mind
heavy ictus heard in English metrical
must be entirely avoided in Latin. If accurate

length is maintained, and accent is always made as
feet
possible, the correct reading of Latin poetry

little diflBculty.
are:
proper length, whether
syllable-
light as
ought to cause
WORDS FREQUENTLY MISPRONOUNCED
90
107.
If,
in fact, proper attention be given to accurate
syllable-length and to such points as elision, slurring, final
m,
the consonant groups of 66 (4), &c., Latin poetry may safely

be read, as far as regards pronunciation, after the manner of
prose.
In any case, Latin verse must be pronounced through-
The
out with the prose accent.
poet's systematised metrical
arrangement of long and short syllables

a quantity-rhythm
will
which ought to be
then emerge in
felt as clearly,
and
appreciated as fully, as accent-rhythm in English.
CHAPTER
XIII

108.
the
In
following
words
false
quantities
common, no matter whether the restored Roman

discarded English pronunciation
(1)
S.mo, basis,
d&bXm,
is
are
or the
in use.
dftmiis, dftrS, fabS,, hSstft, labor,
mantis, mensa, pat6r, quadraginta, quadrtipes.
(In
a is often pronounced as if it were

long in the last two words qudd- is very commonly
made to rime with pod instead of with p&d^. Final
Cf. 9.)
short a must be pronounced distinctly.
these words the

;
(2)
ac, actio, actiim,
amas, amans, audlas, audax, dans,
Sras, fractGm, hortans, Ilia, magniis, mallS, maliimtLs,
Marciis, massa, mavis, maximiis, mensa, mSnebas,
pactttm, radix,
narro,
sal,
(In these words the long a

it
were
1
tactiim, valltim, vasttis.

is
often pronounced as if
short.)
Approximately.
Cf.
4 (a) and 10.

The
91
following groups of words, (3) to (10), are
arranged like the above, according to the particular

vowel (e, e, I, i, 6, o, G, u), frequently mispro-
nounced,
incorrectly shortened, or incorrectly
i.e.
lengthened.
(3)
amarg, bSnS,
giint,
dSiis,
gxpgdio,
divSs, 6a, gis, Snim, So, gques,

Ipsg, IgQ, mgi, mgllor, postgrlor,
illg,
sgnior, siipgrlor, vidg' (or vide).

Igo,
and mgi, are
difficult to
(Words
like ga, dgiis,
pronounce
correctly.
Cf. 9.)
(4)
ames, audlens, bestia, dens,

Ingens,
lecttis, lex,
ensls, festiis, gens, iens,
mens, mensa, metropolis, monens,
mones, nolens, orchestra, penuria, pgrempttts, plebs,

p5tens, prendo, prudens, quotiens, rectiim, regntim,
ren,
rex,
sapiens,
rexi,
sceptriim,
segnis,
securiis,
tecttim, texi, tStlens, vendo, vernSs, vestiglGm, v81ens.
(5)
blcgps, dies, fides, higmg, ibl,
minor, mlniis, mlsgr, nihil,
lit, Ita,
machlna, mlhi,
nisi, pl&s,
prior, quasi,
soclgtas, trlbiis, trlremls, ttbl, vldg (or vide).
such words as
dies,
hlgml,
lit,
practice to pronounce correctly.

(6)
the i requires
(In
much
Cf. 9.)
antiquiis, audis, crispiis, die, digniim,
filltts,
firmiis,
fortiiitiis, frivoliis, ignosco, ins51en8, is (verb), libgrtas,
lictSr, mails,
nonvis,
mavis, millg, mitto, m5neri,
ovilliis,
riidimSnttlm,
nil,
nobis,
princgps, quin, quinqug, radix, rixa,
sic, sis, stilia, tristis, velis
(noun), vesti-
gltim, villa, vis.

(7)
b5n1is,
hlstoria,
bovls,
commodiis,
domi,
d6miim, f6rcgps,
homo, jocus, m6d6, m6ra, movgo, nota,
vldl (with short
) is
occasioDally found (Phaedr., Pers.).
92

novgm, polGs,
positio,
postSrIor, postls, postmSdo,
postpone, proplor, provoco, quomodo,

volens, v61o, volunt.
(8)
bos',
rosS.,
socliis,
Of. 9.
conformo, consciSntia, conscitis,
collaborarg,
consGl, fons, fortiiitiis, frons, hoc, ISborarS, monstro,
mos, noils, nolis, nomlnis, non, nonnS, nonvis, nosco,

oils,
ordo,
orno,
prompttim,
pons, posca,
Qstlftm,
prospSro, prospSriis, rostrftm, SocrStes,
sol, solltudo,
vox.
(9)
ciimiiltis,
dQo,
hiimi,
stimiis, sQis, siltis,
Cf.
(10)
lupQs,
p&gil,
ftrbs,
riidimnt&m,
voliimus, Sxttlto,
gtlttS,.
10, 11.
cur, fmstra, frustfim, fiistls, justfis, jtixtX, lux, nullQs,
nuntiiis, nuntio, plus, rus, rusticGs, struct&in,
sump-
ttim, ultimtts.
109.
Incorrect accent.
other words, the accent
is
In the following and
many
often placed on the antepenult
instead of on the penult

minister, mSglstgr, honestas, hS.rusp6x, proflciscSr, lib^rtas,
siip^rstgs,
f&ciiltas,
paupertas, sScerdos, Lttp^rcJll, tab^mft,
simultas, voluntas, voluptas,
sS,gltt3,, sXt^llSs,
siip^llgx, disclplina, inodorfts, fortiiitiis, &c., &c.
In the following words the accent
is
often placed on the
penult instead of on the antepenult

machlnft, sobrlfts, prlstlntLs, commodiis, v6l&mus, provoco,
1 lO.
quomSdo, p6stmod5, pStugrIm,
Latin words are frequently mispronounced be-
cause of their
'
nihllomlntis, &c., &c.
resemblance to English words.
It is a curious fact
that
bos,
bovU,
is
Cf.
Lat.
very often pronounced bdi, bovU.

sttpSrlSr
and Eng. superior
actio
93
and action sceptrGm with

and miser; sScigtas,
;
sceptre; vestiglttm and vestige; mls6r

society
olivS, olive
sSlena, insolent;
crisptis, crisp
libertas,
liberty;
frivollis,
nSvSm, November; postpone, postpone;

nomlnls, nominal prospgro, prosperous
;
justiis, jt*s<
magistrate
rusticiis, rusticate
hon^stas, honesty
frivolous
in-
c6mmodiis, commodious;
conscliis, conscious;
;
solltudo, solitude
minister, minister; mftglstgr,

fSciiltas,
inodorous; machlna, machine; xc411ens
faculty
inodortis,
= eks-kgl-lens?,
ex-
and a large number of others^

The known pronunciation of the corresponding English word
(not necessarily corresponding in meaning) seems to act as
a trap for the unwary.
cellent; ftmbltlo, ambition;
111.
The
following are a few examples of false quan-
last two or three years, in the House

Commons, the Law Courts, the pulpit, the public platform,
and elsewhere. The Latin phrase, correctly marked, appears
tities heard,
during the
of
on the
ight:
left;
the words as incorrectly pronounced on the
94
(12^
EXERCISES
These
false quantities are
" old " (the "
English
")
95
more commonly heard
in the
pronunciation than in the Roman.
Greater accuracy almost invariably follows the adoption of
Roman
the
of
pronunciation, presumably because the mere act
making the change
necessitates greater attention being
given to quantity.
112.
Roby gives a useful list of words (Grammar,

xc) with the approximately correct phonetic English
spelling.
But they are only approximations after all, and
Vol.
it
is
I,
p.
best not to imitate them.
possible to represent a
word
It
is,
in fact, almost im-
fieri in phonetic English,
\ik.e
unless phonetic symbols are used'.
Professor Postgate also gives a very helpful
with imitated pronunciation.
Primer.
By
to give, in
See
p.
118 of
list
his
of words,
New
Latin
means of the use of different type, he contrives
numerous
phrases, English phonetic equivalents
almost identical with the originals. To those engaged in

teaching the examples are very suggestive.
CHAPTEK XIV
EXERCISES
It is of little avail merely to read through a chapter
new or unusual
acquired only by considerable
dealing with
sounds.
practice,
Accuracy can be
and the reader is
urged to select a few words illustrating each particular

sound (vowel and consonant) and to pronounce them aloud
several times in succession.
probably
suflBce for
1
The
following short
the purpose:
Cf. the artifice suggested at the
end
of 9.
lists will
EXERCISES
98
The vowels.Eead
a: {&b&,
d&tS.,
111.
&Tm&,
cftstrS.,
Smas,
aras, frSgrans,
clarfi,
S.lip^,
pacSt, stratS, plaga,
SmabSt,
clainabS.t.
t6n6r, sSmpgr, pgctSn, SssS, nScn6, bSng, essg, sedes,
e:
tecte, pgdes, b6lle, pStgrS, pSrdSrg, rSpSnte,
pSnderg,
deberS, deleres.
i
tigris,
bibit, sibi, mihi,
pili,
primi, pTri, sibi, crinis,
Imprimis, Imprimis, milltls, minlstri,
tristis, slmllis,
discriminls.
modo, modo,
soror,
consors, tonsor,
odor, cogo, ordo, monstro, nosco,
uomo, uoluo, bono, cognosce, condono,
roboro, coopto.
stimQs, liipus,
Hsus, fructus,
siliis, iirsiis, bftlbiis, fiirfiir,
iructus, fUmiis,
munus, musctis,
bubGltis, ctimGliis,
mQsciiliis.
T^rSs, Sn^x,
gyrtts,
m^rtiis,
IfrS.,
crfpt&, pftpyrtis,
Syria, Byzftntiiim, PyrrhQs, MJtssyli, pyrSmis, gorytfis,
Hylas.
The common
ae
diphthongs.
Read
12
16.
aedes, aestas, caectis, caedes. Paean, minae, mensae,
taedae, aestlmS, praeberS, paed&gogGs, taeniae, AetnS,
saeplsslme.
au
audio, aula, aurft, augiir, auctus, austeriis, caudS, fraus,
gaudgo, laudo, laus, paup6r, rauciis, taurtts.
oe
poena, amoeniis, comoedia, foedarS,
moenlS, oeco-
nomia, Oedipus, OenotrtLs.
Consonantal
lantis
ieci
i (j)
= y. Read
(JanGs),
(jeci),
(ejeci),
iacgo
54
(jac66),
61.
ianiiarltts
(janiiarliis),
igcGr (jgctir), deiSctQs (dejSctiis), eieci
IScQs (jociis), itiba (jQba), iGs (jus), maiSr
(niajSr), peior (pej8r),
maiSstas (majSstas).
EXERCISES
Consonantal u
brSuIs
(v)
= w.
fauor (fSvor),
(brgvis),
(sprevi),
97
priufts
(priviis),
pauo
sSrutis
(pavo),
spreui
(sgrvtts),
uScca
(vaccS), uastiis (vastiis), ugllS (v611g), uia (via), nolo

(volo), u&lntis (vtilniis),
ulx (vix), uox (vox).
Consonants.
c
Read
23.
acSr, acSr, cera, eScidi, c6cidi, cgcini, clbtis,
ctc6r, occldo, occido, facetiis, piscina, spgctes, socigtas,

crticlarS.
ch.
Read
24.
charta, chlmaera, chai^bs, cholgra, lichen,
macblna, schema,
8-
Read
piilchgr,
Snceps,
27.
AchSron, chaos.
gSnSr, gens, IggSrS, plggr, proggnles, sSgSs,
Sglto, Igltfir, magistSr, ggnltis, ggntis, dlgltiis, paglna.
no.
Read
35.
cancSr, cincttis, rancor, princps,
planctiis, concipgrS, concordia, Incido, trttneus, quin-
cunx.
ng.
Read
37

38.
congr&SrS,
fiingi,
iingtila,
longlis, angina', angusttifi,
franggrS,
conjiingerg;
anguis,
sanguis, plnguls, iinguls, lingua, languor.
ph.
Read
42.
phalanx, phantasma, propheta, phrgnesls,
sphinx, graphltim, l^mpha, mgtamorphosls,
n^mpha,
symphonia, sipho.
r.
Read
44.
Roma,
rare, corona,
carcgr, horror, grror, tgrror,

friior,
B.
Read
amor.
46.
(This
is
arma, arbltgr, moror,
marmor,
siirdtis, rorarli,
difficult exercise.)
basilica, otlosfis, pausa, posltio, praesens,
rggens, pars, res, tirbs, trabs, uiilpes, soboles, trans,

'
is
Stowasser, Heinichen, Feyerabend, and others give angina, but this
wrong.
w.
98
EXERCISES
(Avoid the tendency to convert the
pons, dens.
especially s final, into a z.)
a,
Sciplo, sclfintlS, scite,
scilicSt, sceptrtim.
t.
Read
48.
actio, latio, lectio, notatio, ratio, Smicltia,,
Xmbltlo,
cltllis,
comltia, gratis, laetltia, s8ntgntl2,,
tolgrfintia,.
th.
Read
49.
thalamus,
theatr&m, throniis, cathSdrS,
athleta, aether, ethlcS, spSthiila,
mathematica, thorax,
cltharS.
X.
Read
50.
pax,
SxcSlsiis, Sxiilto,
gxactio, tSxtills.
the sound of gz
also
that
rex,
lex,
exills,
Sxpgdio, SxSrcIttis,
6xamn, 8xaspr6,
(Avoid converting the x = ks into
Sxitium,
Sxills
Sxcgls&s
Sxigfttis,
= Sk-si-lis, not eg-zi-lis. Note

Sxcgllens = 6ks-
Sks-k^l-siis
kel-lenss.)
Double consonants. Read

accldr,
SccipSrg,
62-3.
flaccidiis,
floccGs,
adduce, rSddo, rSddltus, addlctio

off Sndo, SffodiQ, officium
saccus
gflfgctiis,
dlfifidlt,
aggrggo, SxaggSro, aggSr,
sGgggro;
alltido,
stlmmfi.m,
commodtim, commungm, commotiim, im-
mobllgm
b611tim, callls, procgUa, cXpUltis
anntiltts,
annalls,
cann&bis,
Innocens
appgllo, pappiis, Sppldiim, appgtitGs, applico

carrtis,
tiissis,
The
addo,
sgrra,
tgssgra
tgrrgstris,
;
horror;
grror,
massa, essg, Sssg,
sagltta, gtittS, attritQs, gGtttir, mittlt.
following pairs of words, though spelt the same, have

and are therefore pronounced differently.
different quantities,
In
many
cases the meanings are entirely different.
cases, too, the

.
origin.
In many
words are of entirely different etymological
EXERCISES
acSr
99
100
EXERCISES
92 (2)
Smabas, Imago,
rSpono, s3,Mt6, Smati,
ftmicae,
pgrsonae, Squestri, s&perbo, monebant, rlifssSnt,

fSciiltas, sftc^rdos.
92 (3)
nattira,
frondosa,
p^rttim, sSrpentls,
92 (4)
92 (5)
cloaca,
sgpultiim,
sgctjris,
bata.
ciultas, fillos,
litigo,
Dc6mbr,
com-
octauiis,
maxime, pr6spr6,
pro-
sci,lp-
comprimo.
m6ngo, mouSo,
l&pldes,
tabella,
fQissSt,
dillgo, diuido, atidles,
c^nstii, rexSras,
sSrint,
92 (6)
formica,
quorundam.
corona, amarg, amata, piidorls, amabar, uldetiir,
uoco,
fiebam, pacatS, audimttr,
legalS,
loricX,
praecalitiis,
ddamas,
paries, cumiilo,
numSro,
montu,
c^cini, uolttcres, cipiSnt, s^lIQ.nt,
fuSrint.
92 (7)
arSa,
libSra,
aiilica,
lm6a,
pOnerS,
quomodo, nautica,
bdrbara, fmproba, subditum, p6ssiimus,
machina, bestia.
92 (8)
fdcgre,
mongor,
/legSrS,
inima, erlmiis,
fl6r6m,
fSrGia, moniilt, r^glmfis, ^picis, m6nlttis, ll,crima,
u6liimiis, regitis.
93(1)
oratores, infelices, uenatores, emgndaui, deleuerunt,
c5mpeglssgnt, SxpSrgisci, dictatores, nonaginta.
93 (2)
rScusaui, rStardaui, salutaui, agasones, obliuisci,

c6cidertint, ciic&rrertint, cScidissSnt, quadraglnta.
93
(3)
elaboro, gl6r!6se, magnlttido, conflt^ri, altitude,

oppgriri, lassltlido, miscueriint, audl^bas.]
93
(4)
ugngrari, pr5flcisci, tgntlissgat, Mltyl^ne, dociierfint, rgsgraul, spgcQlari, Imltari,
monttlssgnt
EXERCISES
93
(5)
101
nolitotg, oratoris, SccusattLs,
bamur, .commutarg,
rgsponderS, hSrta-
rexissemfis, hortabuntftr,
hortaremtir.
93
(6)
coronarg, colorarS, ulatorls, amaulstis, S,inabamiir,
93
(7)
litigator, rexgratls, augiirarS, emlgrarg, insolen-
amaremtis, radim^nttim, Ltipgrcalls, DolabellS.
tgm, annHalS, audlamtii-, audl^ntiir, aggrggarS.
93
(8)
ftnlmarg,
moniifstts,
fiirlosiis,
gr&mma, niimgrarg,
93
93 (10)
(9)
uoltussgm, gpl-
soclarg, homlcida, capltalg.
heredltas, frugalltas, potatio, audfmini, fortlsslmi,

secesslo, postridle,
mSlUsslmae, IndulgSo.
sgueritas, rogatio, rgbelllo,
amaugrant, monemlni,
amauSrlnt, amamlni, pr5batI6, rgmlscgo.
93 (11)
futllltas,
malugro, egregie, bestlolae, relfglo, ac-
cuniGlo, PausAnias, luxuries, tgmflcant.
94 (12)
cgleritas,
monugro,
soclgtas, mls^rlae,
mSnugras,
monugris, rgcupgro, caplmini, rgmouSo.
94 (13)
mirabilg,
tibiclna,
deltidgrS,
fanatlca,
insania,
cogn6scgrg, dlscordia, gmpMciis, Interprgtls.
94 (14)
pgcdnia, c616uxa, domlstlca, pot^ntia, caplssgrg,

boaria, inutilg, amabgris,
94 (15)
materia, mirlflca, malugram, att6iiita, opprimgrg,

c6rp6rga,
94 (16)
amaugram.
mgm6ria,
rgficgrg,
grammtoca, imprlmgrg, obstr^pitiim.
can6iiica, basilica, paribola, mttUgrts,
potugram, quadrupgdg, sacrilgga.
CHAPTEE XV
SELECTED PASSAGES, MARKED FOR READING
In the following passage from Caesar, the marking is
down in the standard
1.
in accordance with the principles laid
work of Dr Bos.
The marking
is
modified to suit English
readers
est
omnis divisa
in partes tres,
quarum unam
Gallia
incolunt Belgae, aliam Aqui-
Celtae,
omnes
legibus
se
ab
Aquitanis
et
Matrona
Ho-
tres,
kwar*
tin'
in-c5-mntBM-gae,&-ll*A-kwltd,-ni,
t^r-tl-a, kw*Ip-s6-riil-
lln-gwa K4l-tae, nos-tra GAl-F

a,p-p6l-l^n-ttir.
H?6m-nes
lin-gwa, in-stl-tti-tis, le-gl-btL
Gil-lo
sin-tgr se dff-fg-rttnt.
sS.-ba-kwi-ta-nis
flti-mSn, a-bel-gis
Ga-rum-nfi
Ma-tro-n*
H6-
dividit.
6t-Se-kw3,-na di-wl-dit.
fortissimi sunt
r6m-nl-iin-for-tis-si-mi stint
kwo
quod
humani-
Bel-gae,
atque
provinciae
longissi-
ta-t pro-vln-cl-ae 16n-gis-sl-
minimeque ad
mercatores saepe com-
m^ iib-siint, ml-nl-me- k w^S-d^-
propterea
cultu
tate
eos
Garumna
Belgis
Sequana
Belgae,
me
Gallos
dififerunt.
rum omnium
a
Galli
Hi
inter
flumen,
nostra
institutis,
appellantur.
lingua,
ipsorum
qui
tertiam,
tani,
lingua
G^l-li* Ss t6m-nls di-wi-s*
Tn-p^r-tes
absunt,
meant, atque
ea,
quae"
animos
eflfeminandos
pertinent,
imiqu*
important,
sunt
ad
prox-
Germanis
prop-te-rg-a
da-kul-tAt-kw^
hu-ma-nl-
os m5r-ca-t6-res sa6-pg
mg-ant,
dt-kwl-a,
dgf-fe-mi-ndn-do
c6m-
kwa-
sd-ni-mos
p^r-tl-ngnt,im-p6r-tant,prok-
sl-mi-kwg
stint
Ggr-ma-nis

qui trans
Rhenum
quibuscum
bellum gerunt.
incolunt,
continenter
103
kwi trans Rhe-nMn-co-lttnt,

kw!-bfis-cung
con-tl-nSn-tgr
b^l-lung g^-riint.
2.
The following passage, from Cicero's Orator, is shown
marked by Ellis. Ellis assumes that his ordinary rules
for pronunciation are known, and here he draws attention
merely to the two points most frequently forgotten, viz.,
syllable length, and the treatment of m final'. (Very slightly
as
modified)
In versu quidem theatra

exclamant, si fuit una
In- ver-su quiden- theatra
tota
tota ex-claman-t
syllaba brevior aut longior.
syl-laba brevior aut lon-gior.
Nee
Nee
multitudo
vero
novit,
nee
tenet; nee illud,
aut
dit,
offendat,
pedes
numeros
quod ofien-
uUos
cur,
aut
intelligit
in
;
et
quo
ta-
men omnium longitudinum

brevitatum in
nee ul-l5s- numeros-
tenet; uec il-lud-,
aut
cur,
quod of-fenin- quo
aut
of-fen-dat, in-tel-ligit
men
una
vero mul-titudo pedes-
novit,
dit,
si fuit
et- ta-
om-niul- lon-gitudinu
sonis, sicut
et- brevitatu in-sonis, sicut
acutarum graviumque vocum, judicium ipsa natura in
actitarung- gravilim-que v6-
auribus nostris collocavit.
auribus- nostris col-locavit.
et
3.
The
cun- judiciu ip-sa natura in
following passage from the Aeneid
(1) as ordinarily printed; (2) as
is
marked by Dr Bos
shown
(slightly
modified); (3) as marked by Ellis (slightly modified); and

(4) " barred " in order to indicate graphically the relative
length of the syllables.

'
Ellis places a
hyphen
after a
consonant that "makes position."
the hyphen after ta- (8tb line) and after vo- (11th line)
is
due
But
to exigencies
of printing, and merely indicate division of words between two lines.
104

(1)
Conticuere omnes, intentique ora teaebant
Inde toro pater Aeneas
sic orsus ab alto

lufandum, Regina, jubes renovare dolorem;
Trojanas ut opes et lamentabile regnum
Eruerint Danai: quaeque ipse miserrima vidi,

Et quorum pars magna fui. Quis, talia fando,
Myrmidonum, Dolopumve, aut duri miles Ulixi,
Temperet a lacrimis ? et jam nox humida caelo
Praecipitat, suadentque cadentia sidera somnos.
(2)
^
Kon-tlk&er^ 6m-nes, ln-t6n-tikw* 6t& tSnebftnt;

In-d6 t6ro pAt8 rae-neas si-kor-sii
In-filn-dtir, reginS,
Troyana
sS,
b^l-to
yubes rSnowarS dolore;
sii-t6pe s5t-lamSn-tabilS reg-nfi"
D&nSi kwaekw^fp-s6 mIser-rimS widi,

Et-kw6rttm pars mag-n3, fui. Kwis-ta-ll-S. f&n-do,
Mj^r-midonQn, DSlopumw^, aut-dtiri milS sG-lj^s-sei,
Tem-pSr6 ta-lA-krImi sSt-yXn-nok sumld& ka^lo
Erufirint
Prae-klpit&t,
swadent-kwe
kftd^n-tlfi sidSrS
s6m-n6s.
(3)
^
Con-ticu^re 6m-nes, in-ten-tique ora ten6ban-t

In-de t6r6 pater Aeneas- sic 6r-sus ab al-to
In-fdn-dur-, regina, jiibes- renovare dol6ren-
Trojanas ut 6pes et- lamen-tabile reg-nu"

Ertierin-t
Ddnai
quaeque
Et- quorum- pr-s mag-na
ip-se mis6r-rima vi-di,

ftii.
Myr-midoDun- Doloplin-ve autT^m-peret a Idcrymis?
et- jAn-
QuIs- talia f4n-d6,

dtiri
miles Ullx-i,
n6x- hiimida co^lo
Praecipitat-, svad^n-tque cad^n-tia sidera s6m-nos-.
105
106
S^cheresse indicates the
4.
first syllable of
each metrical
foot
by printing the contained vowel in Clarendon
The
artifice is useful
La6co6n
dlictus
Neptuno
tyipe.
sorte sac6rdos
SolldmnSs taiirum ing6ntem mact&bat ad aras.
Ecce alitem g6mini a T^nedo tranquflla per alta

(Horresco r^ferfins) imminsis 6rbibus angues
Inclimbunt p61ag6 parit6rque ad
litora
t6ndunt;
P6ctora quorum inter fluctus arrecta juba6que
Sanguineae exsiiperant 6ndas, pars c#tera p6ntum

l^git sinu4ntque immfasa volumine t6rga.
Pone
5.
In his book ffow
to pronounce
Latin, Professor Postgate
some passages of Ancient Latin into a phonetic

based upon the system of the Association Phonetique
transcribes
script
Two
Internationale.
It should
(1)
of these passages are here reproduced.
be noticed that
The vowel values
are in accordance with the symbols
given in the last column of the tabulated scheme
on page
that the word
a, e, &c.,
(3)
Consonantal
(4)
Consonantal u (%) is represented by w.

c (q) is represented by k.
(5)
i (i) is
is nasalised.
represented by
j.
(7)
n before c {q) or g is represented by B.

Trilled r is represented by r.
(8)
Breathed r
(9)
The
(6)
sign
8
'
8.
mean
(2)
This sign
Latin, p. 28.)
|-
(page
is
represented by rh.
represents the vowel sound referred to in

6)'.
found on extant inscriptions. (See How to Pronounce

According to Quintilian, the sign was invented by the
is
Emperor Claudius
for the
purpose of expressing this particular sound.

Final vowels slurred on to
(10)
above the
So
line.
107
vowels are printed
initial
Thus
also nasalised vowels.
omni', kaussa:r".
The
(11)
division of syllables
by a
(12)
omnia
marked where necessary
is
kolligo:, had-ria:tiki:.
shows that the two words are pronounced as
one
I.
stop
full
in-manibiis.
Septimus
mihi
liber
septhmiis
mihi
libier
originum
omni* anti:kwita:tis monl-menta

strium quascumque
in
colligo:
colligo:
nunc
defend!
manibus
in-manibiis
orI:^nust
monumenta
antiquitatis
est
caussarum inlukaussa:r^
cum maxime
strium kwarskiiBkwe deifendi: nuek kim-ma:ks|-me:
inlti:-
conficio
cornfici"
ius
augurium
pontificium
ciuile
tracto:
o:ra:tio:ne:s
j(i:s
auguriiim
pontifikitim
kl:wi:le
trakto:
orationes
multum etiam
Graecis
etiam graekl:s
mult'
more
exercendae
mo:r'
ekserkendae
6tor,
Pythagoreorumque
li:tt6ri:s ti:tor,
py:thagore:o:ruBkw6
litteris
memoriae
memoriae
gratia
gra:tia:
quid
quoque
kwid kwo:kwe
dixerim audierim commemoro uesperi.

di:kser'
audierim
commemoro:
wesperi:
From Cicero de
II.
Phaselus
ille
phase:lus ille
ait
fuisse
ait
fuis-se
quem
kwem
nauium
yidetis
Senectute.
hospites
w|-de:tis hDspite:s
celerrimus
na:wium keler-rimus
neque uUius natantis impetum trabis

nekw' u:Mius natantis impetum trabis
die
die:
108

nequisse
siue
praeterire
palmulis
nekwiis-se praeteriire si:we palmuli:s
opus foret yolare siye linteo

opus foret wolatre si:we linteo
et hoc negat
et
minacis
Hadriatici
ho:k negat mina:kis hadria:tiki
negare
insulasye
litus
Cycladas
nega:re li:tus i:Dsu]a:sw6 kyk-ladas
Rhodumque nobilem horridamque Thraciam

rAoduekwe
Propontida
noibil'
horridaekwe thra:kiam
trucemvje
propontidatrukemwe
Ponticum sinum.
pontikum sinum.
From Catullus.
Note that the distinguishing accents of long and short i

and u are not shown in the second extract. The differences
normally follow the quantity of the vowels, and the accents
are therefore hardly necessary.
The two
extracts should be carefully studied.
Passages from
pronunciation
Feyerabend
Horace and Catullus, with the

marked, are given by Professor
Cicero,
carefully
in his Latin-English Dictionary (pp. xiii
The passages
Feyerabend's method of syllable division

acceptable.
xvi).
are not, however, reproduced here, as Professor

is
now hardly
BIBLIOGRAPHY
1.
Handbook
2.
5.
Primer of Phonetics. H. Sweet. Oxford, 1906.

Elements of Phonetics. W. Rippmann. London, 1899.
Elemente der Phonetik. W. Vietor. Leipsic, 1904.
Lehrbucb der Phonetik. O. Jespersen. Leipsic, 1904.
6.
Petite Phon^tique.
7.
Le Maltre Phon^tique,
"tones"; by Daniel
3.
4.
8.
9.
10.
H. Sweet.
of Phonetics.
Pavil Passy.
Oxford, 1877.
Leipsic, 1912.
Mai Juin
1911.
(Pp.
6081, Chinese
Jones.)
Early English pronunciation. A. J. Ellis. London, 1869.

Visible Speech.
Bell.
London, 1867.
Outlines of Latin Phonetics.
Niedermann.
Translation by
Strong and Stewart. London, 1910.
11.
Quantitative pronunciation of Latin.
12.
The Soman pronunciation
13.
The Boman pronunciation
A.
J. EUis.
of Latin.
J. B. Scheier.
of Latin.
F. E. Lord.
London, 1874.
Notre Dame,
Indiana, 1894.
Boston, U.S.A.,
1895.
14.
Pronunciation of Latin in the Augustan period.
Camb.
Phil.
Soc., 1905.
15.
Restored pronunciation of Greek and Latin. Arnold and Conway.

Cambridge, 1906.
16.
How
17.
Thôrie generale de I'Accentuation Latine.
18.
Petit Traits de Prononciation Latine.
19.
Livre de Lecture Latine.
20.
Traits ifil^mentaire de Prononciation Latine.
to
pronounce Latin.
J. P. Postgate.
London, 1907.
Weil et Benloew.
Paris, 1855.
resse.
Paris, 1903.
A. Bos.
A. Bos.
Paris, 1893.
Paris, 1897.
Ariatide S^che-
BIBLIOGRAPHY
110
21.
Ausspraohe, Vokalismus und Betonung der lateinischen Spraohe
22.
Die Aussprache des Latein. Emil Seelmann. Heilbronn, 1885.

Hulfsbuchlein fiir die Aussprache der lateinischen Vokale in
23.
Leipaic, 1870.
Corsseu.
(2 vols.).
Anton Marx.
positionslangen Silben.
1901,
now a good
(Last edition,
Berlin.
deal out of date.)
24.
Lautlehre der lateinischen Sprache.
25.
Handbuch der
lateinischen Laut-
F. Stolz.
Leipsic, 1894.
und Formenlehre.
F.
Sommer.
Heidelberg, 1902.
Leipsic, 1855
26.
Grammatici Latini,
27.
Grundriss der Romanischen Philologie.
1888
28.
29.
ed. Keil.
1880.
Grober.
Strassburg,
ff.
The Latin Language. W. M. Lindsay. Oxford, 1894.

The Latin Language, especially pp. 36 72, on hidden quantities.
C. E. Bennett.
Boston, U.S.A., 1907.
30.
Latin Grammar, Part
31.
School Latin Grammar.
32.
33.
Public School Latin Grammar.

The New Latin Primer. J.
34.
Latin Grammar.
35.
Grundriss der vergleichenden Grammatik.
I.
Roby.
Roby.
London.
London.
Kennedy.
P.
London.
Postgate and 0. A. Vince.
London.
sic.
Vol.
Hale and Buck.
Boston, U.S.A., 1903.
K. Brugmann.
Leip-
1897.
I,
Grammatik der
39.
lateinischen Sprache (2 vols.).

Hanover.
Precis de grammaire comparee du Grec et du Latin. Victor
Henry. Paris, 1894.
Grammaire comparee du Grec et du Latin, Vol. i. O. Riemann
Paris, 1897.
et H. Goelzer.
Manual of Latin Prosody. (Seventh edition.) W. Ramsay.
40.
Traits de Versification Latine.
41.
Nouvelle Prosodie Latine.
42.
De
36.
Ausfiihrliche
New
37.
38.
Edition.
Kiihner.
London.
L. Quicherat.
L. Quicherat.
Paris, 1882.
Paris,
1903.
43.
St Petersburg and Leipsic.

Corpus Inscriptionum Latinarum. Berlin, 1863 flf.
44.
re Metrica.
L. Mueller.
Latin Dictionary for Schools.
Lewis.
Oxford, 1901.
than Lewis and Short for hidden quantities.)
1894.
(Better
111
BIBLIOGRAPHY
45.
Lateinisch-Deutsches Schul- und Handworterbuch.
Stowasser.
Leipsic, 1910.
46.
Latin-English Dictionary.
47.
Kleines Lateinisch-Deutsches Schulworterbuch.

vised
48.
by Blase and Reeb.
Heinichen
re-
of
Brugmann's Com-
Classical Review, Vol. xix, 1905, " Vendryes,
and the Ancients,
Grammar.
on Greek Accents."
50.
Berlin, 1912.
Leipsic, 1911.
Classical Review, Vol. xiii, 1899,
parative
49.
K. Feyerabend.
J.
Review
P. Postgate.
J. P. Postgate.
Journal of the British Academy, Vol.

Research.''
lii,
1908, " Flaws in Classical
J. P. Postgate.
No.
" Latin syllabification
51.
Classical Philology, Vol.
52.
Walter Dennison.
Classical Philology, Vol. in, No. 1, "Accent in Latin."
l.
1,
evidence
of inscriptions."
J. P.
Post-
gate.
53.
Hermathena, 1908, "'Sprechtempo' or Phonetic Law?"
54.
Papers on Hidden Quantities. See p. 59, footnote. Professor

Buck's paper in the Classical Review for June 1913 is of
Professor Sonnenschein writes a short final
special interest.
See also
reply in the Classical Review for August 1913.
C. Exon.
vol. 29, supra.
OAMBBIDGE: PRINTED BI JOHN CLAY,
M.A.
AT THE nNrVERSrTT PRESS.

Quantity and Accent in The Pronunciation of Latin - F. W. Westaway (1913) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Quantity and Accent in The Pronunciation of Latin - F. W. Westaway (1913) PDF

Uploaded by

Copyright:

Available Formats

<^%CA,^^

CORNELL UNIVERSITY LIBRARY

3 1924 064 122 660

Cornell University Library.

There are no known copyright

the United States on the use of the

BOUGHT WITH THE INCOME

SAGE ENDOWMENT FUND

QUANTITY AND ACCENT

CAMBRIDGE UNIVERSITY PRESS

ASHER AND CO.

All rights reserved

QUANTITY AND ACCENT

the University Press

"Si nimius videor, seraque coronide longus

Terque quaterque mihi

correspondence in The Times during the past year

of that older race of classical scholars

let alone in their

easy-going pronunciation of Latin.

was their custom

who were ever clamant

anyone in their hearing

such words as antlquus, amicus, fortuna, or imperdtor, with

with the accent on the ante-

penult, the offending person was promptly cast without the

be curiously unconscious of their own

they would shorten the long vowels in

non, vox, nil, lex, lectus^, rex, rectum

they would lengthen tKe short vowel in the accented syllable

they frequently accented the antepenult instead

of the penult in viintster, facultas, liMHas, exdllens; and

other analogous words.

and sin^ with

occasionally hear of poor "

(mindr) being called upon, before he makes another attempt to

struggle through a few lines from the "Mete(r)m<^rphe(r)ses"

does not the old-time K.C., engrossed as he

get rid of a troublesome case by putting forward a "nolly

of this old school pleads for a retention of

old " or " English " pronunciation because, they say,

would save trouble in teaching.

Is not the poor British

school-boy worried enough already'

Let him, then, pro-

nounce Latin as he pronounces English.

pronounced fide and sin^ after the English pattern, both

and such words as antiquus,

amicus, casttgo, minister, magtster, libSrtas, facultas, honistas,

pronunciation of the analogous English words antiquated,

desire to adhere to the

old or English pronunciation have realised not only

knowledge of Latin quantities, but

that, in the adoption of the restored or

they are necessarily in danger of being found out.

Accushave been almost from the cradle to the use

The wise schoolmaster does

not, of course, worry his pupils with rules

about the minutiae of Latin pronunciation.

to adopt a system in which

pronunciation, they hesitate

demand such scrupulous

Opposition to reform sometimes comes from the most