MICROFILMED 1991 
 
 COLUMBIA LNIVERSrr\ LIBRARIES/NEW YORK 
 
 as pan oi the 
 
 ■"Foundations ot Western Ci\'ilization PreservaUon Proiecr' 
 
 »,/ 
 
 Funded b\ the 
 
 ATIONAL ENDOWMENT FOR THE HLAlANITIES 
 
 Reproductions may not be made without permission from 
 
 Cokmibia University Librar}' 
 
COPYRIGHT STATEMENT 
 
 The copyright law of the United Slates - Title 17, United 
 States Code -- concerns the making of photocopies or other 
 reproductions of copyrighted material.., 
 
 4 
 
 Columbia University Library reserves the right to refuse to 
 accept a copy order if, in its judgement, fulfillment of the order 
 would involve violation of the copyright law^ 
 
AUTHOR: 
 
 NIEDERMANN, MAX 
 
 TITLE: 
 
 OUTLINES OF LATIN 
 PHONETICS 
 
 PLA CE : 
 
 LONDON 
 
 DATE: 
 
 1910 
 
COLUMBIA UNIVERSITY LIBRARIES 
 PRESERVATION DEPARTMENT 
 
 Master Negative # 
 
 I I / » 
 
 i > I . I 
 
 )( 
 
 
 i f ■ \ 
 
 V : A 
 
 
 Resfnctions on Use: 
 
 Original Material as Filmed - Existing Bibliographic Record 
 
 877.15 
 
 W553 
 
 I 
 
 NiedGrnann, :'ax, 1S74- j^/j^ 
 
 Outline:; of Latin plionotics, ed. by V.. A, | 
 
 otro>ip, n.nd II. :;,tc77art. London, Rc;nt1':dpG, 191C, 
 
 vii, 107, j-1^ p. diap,rs. 19 cr>. 
 
 
 .*-. ,f*^ 
 
 t I 1 .' IV 1 ^j 1 /Ij Cf 
 
 _) ft'^i rv^i 
 
 k I ! ^ »►„ • -'T' 
 
 
 ■*--^^ • 
 
 *^ r-* 
 
 ARCH pi 
 
 TECHNICAL MICROFORM DATA 
 
 REDUCTION RATIO: 
 IB IIB _ 
 
 i INITIALS___i:iS^__. 
 
 .AilONS. INC WOODBRIDGE. CT 
 
 Ik 
 
tBy^.iivr*'-!':^'^^^ 1 
 
 C 
 
 Association for information and Image {Management 
 
 llOOWayne Avenue. Suite 1100 
 Silver Spring, Maryland 20910 
 
 301/587-8202 
 
 v^ 
 
 Centimeter 
 
 1 2 3 
 
 Liil 
 
 lllllllllllllllllllllllllllll 
 
 ITT 
 
 llillllillllllllillil 
 
 iiilniiliiiiliiiiliil 
 
 TTT 
 
 7 8 9 
 
 ilmijmiJLi^^ 
 
 T 1 I 
 
 TTI 
 
 10 
 
 11 12 13 14 15 mm 
 
 iiliiiiliiiiliiiiliiiiliiiiliiiiliiiiliiii 
 
 Inches 
 
 .0 
 
 LI 
 
 1.25 
 
 ■ 56 
 
 Ijo 
 
 2.8 
 3.2 
 
 36 
 4.0 
 
 1.4 
 
 2.5 
 
 2.2 
 
 2.0 
 
 1.8 
 
 1.6 
 
 MFINUFflCTURED TO fillM STfiNDflRDS 
 BY APPLIED IMRGE. INC. 
 

Columbia (HniUf rsftj) 
 
 intI)fCitpofBrtnl0rk 
 
 THE LIBRARIES 
 
 /I 
 
 » 
 
 OUTLINES OF LATIN PHONETICS 
 
 i 9 
 

 OUTLINES 
 
 OF 
 
 t 
 
 *#^- 
 
 LATIN PHONETICS 
 
 BY 
 
 PROFESSOR MAX NIEDERMANN 
 
 «^% 
 
 EDITED BY 
 
 PROFESSOR H. A. STRONG, M.A., LL.D., 
 
 EMERITUS PROFESSOR OF LATIN, LIVERPOOL UNIVERSITY 
 
 AND 
 
 H. STEWART, B.A. 
 
 LECTURER IN CLASSICS, LIVERPOOL UNIVERSITY 
 
 . ^ '^ 
 
 LONDON: 
 GEORGE ROUTLEDGE '8: SONS;tlMlTED 
 NEW YORK: E. P. DUITON & C'O. " * ^ 
 
 1910 
 
 <• • * 
 
•^./ 
 
 i> 
 
 ^ 
 
 
 -5 
 
 
 ^11. IS 
 K 6'5'3 
 
 \ « 
 
 • « * 
 
 « • • • 
 *% • t 
 
 • » . 
 
 I • • 
 
 t > • 
 
 • • III 
 
 • t « 
 
 f > • 
 
 t • 
 
 t 
 « * t • « 
 « I ' I • 
 
 « 
 
 • • • » 
 
 , . . I 
 
 • « 
 
 tit « 
 
 INTRODUCTION. 
 
 It has long been recognised that the methods of teaching 
 Latin grammar in vogue in England are marked by an 
 absence of scientific accuracy and by a neglect of philo- 
 logical principles. The conditions that Professor Meillet 
 in his foreword to this book describes as existing in France 
 obtain in this country as well : ' Separated from all modern 
 conceptions, and still scholastic in its essence and formation, 
 grammatical instruction has become a dead element; it 
 is confined to facts unconnected by any ideas or connected 
 only by ideas which are now obsolete and conflicting with 
 everything else learnt by the pupil. Thus every day 
 opinion becomes more hostile to grammar, and every year 
 fewer hours are assigned to a subject the barren nature 
 of which is felt by every one. Yet,' as the same critic 
 continues, ' it might be thought that a science the essen- 
 tial object of which is to study the principal means of 
 expression of human thought might deserve the attention 
 of schoolboys, and be capable, if properly taught, of 
 awakening their lively interest.' 
 
 In the hope, then, of filling to some extent an admitted 
 blank in the public school curriculum, the editors have 
 translated this little book, which has met with a large 
 measure of favour on the Continent. It may prove use- 
 ful in University courses as well as in schools. It was 
 
\ 
 
 ^ LATIN PHONETICS 
 
 intended, however, not for students, but for schoolboys 
 between the ages of fifteen and eighteen with a sufficient 
 grounding in the elements of Latin. The author has 
 omitted references to Greek, believing that a large 
 majority of such boys know nothing, or next to nothing, 
 of that language ; and, as Professor MeiUet remarks with 
 justice, this has also definite advantages from a purely 
 scientific point of view: Latin has closer connections 
 with Oscan and Umbrian and the Celtic dialects than 
 
 with Greek. 
 
 The translators have compiled an index, but have 
 refrained, except in a few instances, from inserting notes, 
 not only from reluctance to burden the outline with 
 additional matter, but chiefly in the conviction that 
 further elucidation and comment should be left to the 
 teacher. 
 
 * before a word indicates a hypothetical form restored 
 
 by induction. 
 
 > means 'developed into'; U, e > i means 'e 
 
 developed into i.' 
 
 O.L. = Old Latin. 
 
 ) ( means * contrasted with.' 
 
 A small dash before or after, or both before and after 
 a letter or a combination of letters, marks an initial, 
 final, or medial sound; e.g., 5- = initial 5, -s = final 5, -s- = 
 
 medial s. . 
 
 Since Latin presents numerous grammatical coincidences 
 with Greek, Celtic, Slavonic and several other languages, 
 we must conclude that all these languages develop from 
 
 ' Mi 
 
 # 
 
 « 
 
 f 
 
 INTRODUCTION 
 
 Vll 
 
 a common source, that they continue a prehistoric common 
 language just as French, Itahan, Spanish, etc.. continue 
 Utui. rh.s prelustor,c language is known as the Indo- 
 European ; ,ts various transformations, represented by the 
 languages to which we have just referred, are called Indo- 
 iiiuropean languages. 
 
 after the termination .s cut off. The termination is the 
 variable element which serves to indicate the part played 
 by the word m the sentence. Thus due is the stem of 
 du^, manu- of manus, ama- of amare, and so on The 
 division into stem and termination in Latin is often 
 difficult, and in several cases quite impossible without 
 the help of Its cognate languages. Sometimes the stem 
 cannot be reduced further, as. for example. d,l: in du., 
 but general^ it can be split up into two quite distinct 
 factors, each of which is found in other combinations. 
 Consider, for example, eantu-, the stem of cantus, -us. 
 The ebment can- is seen also in cano, canor, cantor, etc 
 and the element -tu- in gustus, -us; partus, -us; sumptu^, 
 -us etc Similarly the analysis of rube-, the stem of 
 rubere shows a factor rUb- which is contained also in 
 ruber, rubor, rubia ('madder,' a plant from whose roots 
 IS made a red dye), and a factor -e- found in cafeye 
 latere, nitere, etc. Of these two constituent elements of 
 the stem the first is called the root, the second the suffix 
 Ihus a word consists normally of three parts-root, suffix 
 and termination. 1 ' 
 
 1 [From * on is translated from the Frfinrh v^u^ • ^ 
 oat«d throughout by «,uare b Jke,^ ] ^"°"''' ""**' "« '-"l'- 
 
 >'*' 
 
 '■^1 
 

 LATIN PHONETICS. 
 
 # 
 
 * 
 
 PBELIMINARY OBSERVATIONS. 
 
 § 1. Phonetics is the science dealing with the sounds 
 which constitute language. By sounds are meant all the 
 impressions that are made on our organs of hearing 
 ^ determined by the modifications which the organs of 
 speech produce on the current of breath issuing from the 
 lungs. These impressions may be investigated from two 
 points of view : (1) from that of their physiological pro- 
 duction, and (2) from that of their historical development. 
 Indeed, sounds depend for their development on space 
 H and time. The transformations which they undergo occur 
 unconsciously, i.e., independently of the will or reflection 
 of the speaker. They are brought about by imperceptible 
 gradations, and according to unchangeable laws. That is to 
 say, all sounds, or groups of sounds, if exposed to identical 
 conditions, develop with absolute regularity. Exceptions 
 11^ to this rule, which is the fundamental principle of the 
 linguistic method, are merely apparent. They are mostly 
 to be explained as being the result of disturbances 
 in the normal course of phonetic laws brought about 
 by a psychological cause, which we call analogy. For 
 example, in Latin, in an open medial syllable {i.e., where 
 0^ the vowel is followed by a single consonant only), a 
 short (T appears unchanged in such words as comeTdo, 
 
2 LATIN PHONETICS 
 
 app«^to, obt<?go, diss<9C0, subv«9nio ; while it becomes i in such 
 words as praesideo ( ) ( s^deo), constiti ( ) { sUti), colligo 
 ( ) ( lego), ausplcis (genitive of ausp^x), sustmeo ( ) ( t«9neo). 
 But we must be careful not to regard this difference of 
 treatment as a capricious trick of language. As a matter 
 of fact, the transition of e into ^ took place in all the 
 words mentioned ; but analogy has restored the original 
 € in comedo, app^to, obt^go, diss^co, subv^nio, in order to 
 accentuate their affinity with 6do, p6to, etc. Thus analogy 
 acts with a levelling tendency on words related by gram- 
 matical function or signification. 
 
 A 
 
 
 
 f! 
 
 GENERAL THEORY OF SOUNDS 
 
 GENERAL THEORY OF SOUNDS. 
 
 § 2. Spoken language is produced by the current of air 
 issuing from the mouth in the shape of breath. The 
 breath, expelled from the lungs, passes first through the 
 windpipe and then into the larynx. The larynx is formed 
 of several cartilages, between which are stretched two 
 membranes— the vocal chords. Between these again lies a 
 slit called the glottis. If the current of breath finds the 
 glottis closed, it forces its way out by a series of jerks 
 which set the vocal chords in vibration. On the other 
 hand, if it finds the glottis open, it passes freely through, 
 and sets up no movement of the vocal chords. In the 
 former case the sound produced is said to be voiced ; in 
 the latter case, breathed. A simple method of testing to 
 which of these two categories any given sound belongs is 
 to close the ears while uttering the sound. If you hear 
 a buzzing noise in the head, the sound produced is voiced ; 
 if no buzzing sound is heard, it is breathed. After passing 
 through the glottis the breath enters the pharynx, from 
 which it escapes through the mouth or the nose, or 
 through both exits simultaneously, according as the soft 
 palate, a membraneous partition between the cavity of 
 the mouth and that of the nose, is raised, depressed, or 
 held in balance in an intermediate position by its muscles. 
 The mouth plays a double part in sound production. 
 Sometimes it serves merely as a sounding-board to the 
 sound produced in the larynx ; at other times, by a series 
 of stoppages, or narrowing processes, it puts various 
 
Jt 
 
 4 LATIN PHONETICS 
 
 obstacles in the way of the breath. Consequently sounds 
 are divided into (a) those in the utterance of which 
 the mouth cavity maintains unchanged the position it 
 has once adopted, and {b) those which depend entirely 
 or mainly on a closing or opening movement. The 
 former are called vowels ; the latter consonants. It is 
 important, however, to notice carefully that one cannot 
 draw an absolute line between these two classes of 
 sounds [e.g.i the sonant use of consonants /, tw, n (also 
 ng^ i.e., «^), r, z, represented as /, ^, etc., spelt, e.g., 
 TdXtle, Birmingham, writt^w, can ^ive, as\ f is not 
 represented in English, but is common in German, e.g., 
 * butter.'] On the other hand 2, ou, ii are true vowels in 
 the French words je lie, je tue, je loue; but as they 
 admit of a distinct raising of the tongue, they can act 
 also as movements of closing and opening, that is to 
 say as consonants ; compare nous lions, nous tuons, nous 
 louons, where, in modern French, each word is pronounced 
 as a monosyllable. With this reservation we may for 
 convenience maintain the accepted terms vowel and 
 consonant. 
 
 § 3. Starting with the neutral vowel a, which is pro- 
 nounced with the tongue lying loosely on the base of the 
 mouth cavity, we may divide the vowels into front and 
 hack vowels. This distinction is based upon whether the 
 tongue lies bunched in the front or the back part of 
 the mouth. Another possible division may be drawn 
 dependent on the vertical distance between the top of 
 the arched part of the tongue and the roof of the hard 
 palate. In this way we distinguish between low and high 
 vowels, more commonly known as open and close vowels.^ 
 
 1 Cf. Giles, § 79. 
 
 ^ 
 
 I * 
 
 ^K 
 
 <> 
 
 GENERAL THEORY OF SOUNDS 5 
 
 Two vowels of different timbre or acoustic quality, 
 uttered without break in the same syllable, constitute 
 a diphthong. 
 
 Diagram of the Organs of Speech (after Victor). 
 
 
 m 
 
 (a) nasal cavity. 
 (6) hard palate, 
 (c) soft palate, 
 (rf) mouth cavity. 
 {e) tongue. 
 (/) pharynx. 
 {g) hyoid bone. 
 (h) epiglottis, 
 (i) rima glottidia. 
 {k) vocal chord. 
 (0 thyroid cartilage, 
 (m) laryngal cavity. 
 (n) tracheal rings. 
 (o) air tube. 
 [p) oesophagus. 
 
 >\;' 
 
 § 4. Consonants fall into two main categories : close 
 or stops, and narrow. The first are characterised by a 
 complete stoppage, the latter by a mere narrowing of 
 the air passage. The stops are instantaneous explosive 
 sounds ; the narrow, on the other hand, can be maintained 
 for a more or less long time. Hence the stops are some- 
 times called 'momentary' and the narrow 'continuous' 
 sounds. The narrow are divided into four categories, 
 which are named after the manner in which the air 
 
p=^5«y|jii.s '» •■".-. -> .^r---! 
 
 1/ 
 
 6 LATIN PHONETICS 
 
 escapes in the process of their production : Fricatives or 
 Spirants, Trilled Consonants, Lateral or Divided Con- 
 sonants (these last two are often spoken of by the common 
 name of Liquids)^ and Nasals. 
 
 In the production of spirants the current of air finds 
 the breath passage of the mouth narrowed, and produces 
 a rubbing sound. 
 
 The trilled consonants are determined by the rapid 
 vibration against the teeth roots of an elastic organ (the 
 uvula or tip of the tongue), which is displaced by the 
 breath, and then returns to its former position [or by 
 vibration of the flexible parts of the mouth between 
 themselves, e.g.^ the lips]. 
 
 Divided consonants are formed when the tip or back of 
 the tongue is pressed against the central line of the hard 
 palate, from the teeth to the soft palate, and blocks the 
 path for the breath, which escapes sideways. 
 
 Finally, the nasals are produced when the mouth is 
 closed, but the soft palate lowered, so that the air can 
 pass freely through the nose. 
 
 Another division of the consonants may be made accord- 
 ing to the place in the mouth where the obstacle arises, 
 which conditions the sound characteristic of each. 
 
 [Divided thus there are seven main classes : — 
 
 n' 
 
 GENERAL THEORY OF SOUNDS 7 
 
 and the roots of the upper teeth, e.g.y i as in temper (stop) ; 
 s as in summer (spirant) ; n as in nowhere (nasal). 
 
 (5) Interdental, formed by the tip of the tongue and the 
 tips of the front teeth, e.g., th as in then, thin (spirants). 
 
 (6) Labiodental, formed by the lower lip and upper 
 teeth, e.g., f And v as in /ally veal (spirants). 
 
 (7) Labial, formed by the upper and lower lips, e.g.,p, b, 
 w, m (and sometimes r when used alone as an ejaculation), 
 as in pile, beer (stops); woman, man (nasals).] 
 
 4' 
 
 /> 
 
 (1) Velar, back or guttural, formed by the root of the 
 tongue and the soft palate (velum), e.g., ch as in Scotch 
 loch (spirant) ; ng or i9 as in being (nasal). 
 
 (2) Palatal or front, formed by the middle of the tongue 
 and the hard palate, e.g., k before c or i in such words as 
 Kent, kindred (stop). 
 
 (3) Cerebral, formed by the blade of the tongue and the 
 front of the hard palate, e.g.^ s = shj z = zh, as in daze 
 (spirant). 
 
 (4) Dental (alveolar), formed by the tip of the tongue 
 
 *v 
 
 f^ 
 
'■'"'-,' '* '^ 
 
 
 t i=»*^^'a-S';;;^^ TV, f w^^'t»:» 
 
 ^*> 
 
 8 
 
 LATIN PHONETICS 
 
 CLASSIFICATION OF THE LATIN SOUNDS. 
 
 § 5. The Latin vowels may, according to the principles 
 laid down above, be represented by the following scheme, 
 which shows the different positions of the tongue : — 
 
 « 03 
 
 o o 
 
 03 3 
 
 w a 
 
 o 
 
 Tongue 
 in front of mouth. in back of mouth, 
 least > i u 
 
 I U 
 
 e 6 
 
 6 ^ o 
 most > a 
 
 Ncfte. — The vowel y, n©t mentioned in the scheme, occurs 
 only in Greek loanwords (see § 6, note iv. 3). 
 
 There were four diphthongs in classical Latin, viz. : — 
 eu (very rare, see § 22), ae, au, oe. 
 
 § 6. In the following Table the consonants are classified 
 from the double standpoint of the form and place of 
 their articulation : — 
 
 ^. 
 
 v*> 
 
 »» 
 
 
 ^f 
 
 ^ 
 
 h 
 
 CLASSIFICATION OF THE LATIN SOUNDS 9 
 
 Form of 
 Articulation. 
 
 d 
 
 Cm O 
 
 c3 o 
 < 
 
 ■ 
 
 03 
 
 CO 
 
 '3 
 
 a 
 
 Q 
 
 • 
 
 2 
 
 Stops. 
 
 ■ 
 
 voiced. 
 
 b 
 
 d 
 
 g 
 
 breathed. 
 
 P 
 
 t 
 
 c, k, q 
 
 Narrow. 
 
 CO 
 
 02 
 
 voiced. 
 
 
 
 
 breathed. 
 
 f 
 
 s 
 
 
 • 
 
 H 
 
 voiced. 
 
 
 r 
 
 
 breathed. 
 
 
 
 
 Q 
 
 voiced. 
 
 
 
 1 
 
 breathed. 
 
 
 
 
 Nasals. 
 
 voiced. 
 
 m 
 
 n 
 
 Id 
 
 breathed. 
 
 
 
 
 [For fuller table see Giles, table opposite, p. 90.] 
 

 *^ 
 
 fs 
 
 10 
 
 LATIN PHONETICS 
 
 CLASSIFICATION OF THE LATIN SOUNDS 11 
 
 Notes. 
 
 (i.) The place of articulation of the guttural stops is not 
 fixed. The tongue touches the palate more or less back 
 according to the acoustic quality of the following vowel. 
 Before e and i the position is at the hard palate ; before a 
 at the middle of the palate, i.e., half-way between the 
 hard and soft palates ; before o and u at the back part of 
 palate, i.e., the soft palate or velum. It is probable that 
 in O.L. the use of the three symbols c, k, q, corresponded 
 to these three positions. Indeed, it seems that originally 
 c was as a rule written before i and e {c'lvis, cima ), k before 
 a (kcipnt, karus), q before o and u (qSm^s, qura). Gradually, 
 however, the use of c became general ; k disappeared and 
 left but a few traces in certain abbreviations, e.g.,'K = Kaeso 
 (proper name), K or KAL^calendae, KA=capimlis, KK = cast- 
 rorum, etc. ; while q was retained only in the combination qu, 
 which denoted the breathed labialised guttural in words 
 like qui, quae, quod (§ 39). Latin recognised also two kinds 
 of the narrow divided I, which was articulated sometimes 
 at the root of the upper front teeth (palatal I) ; sometimes 
 at the soft or velar palate (velar I). The / was palatal, 
 (rt) when initial ; (6) when medial before i ; (c) in cases 
 of doubled II. It was velar {a) when final ; (6) when 
 medial before e, a, o, u, and before a consonant. 
 
 (ii.) The sound represented by Latin / was originally 
 formed by pressing the lips together. But at a relatively 
 early period this bilabial ^ changed into a labiodental /, 
 which, as in English, French, and German was produced 
 by the pressure exercised by the upper front teeth on the 
 lower lip. 
 
 (iii. ) 10 denotes the guttural nasal, for which the Latin 
 alphabet like English, French, and German possessed no 
 special symbol. Before a guttural stop this sound was 
 written n, as in angulus and ajiceps ; before a nasal, g as 
 in dignus (§ 69). The poet Accius (170-94 B.C.), proposed to 
 employ g before gutturals as well, as was done in Greek. 
 Thus instead of angulus, anceps, ancora, etc., he would 
 
 I 
 
 » 
 
 ^i9 
 
 ?> 
 
 vii 
 
 >* I # 
 
 ^ [See Edmonds, pp. 77, 78. The bilabial / was represented in early 
 Latin by /A, e.g., FHEFHAKED, the old form of fecit.] 
 
 ^ 
 
 have written aggulus, agcepSf agcora. His suggestion, how- 
 ever, was not adopted. 
 
 (iv^) In the scheme given above the following letters are 
 omitted : — 
 
 {!) h, which was mute in classical times as in modern 
 French (§ 44). 
 
 (2) j and v, which were semi- vowels, at least as late as 
 the first century A.D., i.e., consonantal forms of the vowels 
 i and u. Their sounds were equivalent to those heard in 
 i and u respectively in the French ^ncrre (pronounced pyer 
 and 6chouer (pronounced echwe). The use of the letters j 
 and V dates from modern times. Their introduction is due 
 to the scholars of the Renaissance. In this book these 
 convenient symbols will be employed to represent initial 
 consonantal i and u. The Latin alphabet, however, knew 
 only i and 21. 
 
 (3) The voiced dental spirant ~, which was unknown in 
 genuine Latin words, of. Quintilian, Iiistit. Oral., xii. 10, 
 28 ; namque est ipsis statim sonis durior (i.e., Latina 
 facundia) quando et jucundissimas ex Graecis litteras non 
 habemus, vocalem alteram, alteram consonantem, quibus 
 nuUae apud eos dulcius spirant, quas mutuari solemus 
 quotiens illorum nominibus utimur — quod cum contingit, 
 nescio quomodo velut hilarior protinus renidet oratio, ut in 
 zephyris et zophoris : quae si nostris litteris scribantur, 
 surdum quiddam et barbarum efficient. 
 
 (4) The breathed aspirate stops ch, ph, th, originally 
 restricted to Greek loanwords borrowed at a compara- 
 tively late period, such as chdrus, mdchina, phiUrum (a 
 love charm), romphaea (a broad-sword), thesaurus, cUhdra. 
 Under the influence of these aspiration was introduced, 
 it is true, from the end of the second century B.C., in a 
 certain number of words of genuine Latin origin, as in 
 pulcher (the pronunciation pulcer was obsolete in Cicero's 
 time, as he himself informs us, cf. Orator, 48, § 160) ; s^ul- 
 chrum (a pronunciation which in Cicero's time was not 
 common, and one which Cicero personally rejects (cf. 
 Orator, loo. cit.), limpha (spring water), sulphur, etc. 
 (§ 35). 
 
 (v.) The difference which manifests itself in English in 
 the names given to the letters denoting the consonants (/, 
 
 ; 
 
* I * 
 
 12 LATIN PHONETICS 
 
 I, m, n, r, 5, being called ef, elly emm, enn, ar, ess, while on 
 the other hand, 6, c, rf, gr, k, p, q, t are called 6e«, cee^ dee^ 
 gee^ kay, pee, kew, tee) goes back to ancient times with the 
 reservation that until the second half of the fourth century 
 A.D. the letters of the first group named /, I, m, etc., were 
 pronounced without any supporting vowel. Thus up to 
 this time they were * litterae per se nominativae sive quae 
 per se prolatae nomen suum ostendunt ' (Charisius, i. p. 9, 
 4 K and Priscian, ii. p. 8, 10 K) just like a, g, i, o, u. 
 Hence the Roman grammarians called them semivocales 
 in opposition to b, c, d, g, etc., which were called mutae ; 
 • quod per se sine adminiculo vocalium non possunt enun- 
 tiari ' {DiomedeSy i. p. 423, 24 K). 
 
 ? • 
 
 >' 
 
 I 
 
 LATIN ACCENTUATION 
 
 13 
 
 LATIN ACCENTUATION. 
 
 § 7. In the sound groups which we call words, one of 
 the syllables in each word may be brought into promin- 
 ence either by strong muscular effort or by a more or less 
 marked elevation of the voice. 
 
 The degree of prominence obtained by special energy 
 in articulation forms the stress accent, or, shortly, * the 
 accent ' ; that which is due to the change from a lower to 
 a higher musical note makes the pitch accent or tone. 
 
 Now, accent and tone may exist simultaneously in one 
 and the same language. This was the case in pre-literary 
 Latin. Independently of the alternations, inherited from 
 the Indo-European language, of high and low syllables, a 
 sharply-marked stress accent had begun to mark the first 
 syllable of words. But at the beginning of the literary 
 period this stress accent disappeared, as it was ill-suited 
 to the genius of the Latin language, in which the quan- 
 tity of the syllables was strictly marked. From the 
 second century B.C. down to the fourth century a.d., 
 Latin possessed merely a pitch accent or tone, the place 
 of which was regulated by the quantity of the penultimate 
 syllable. 
 
 There are no traces left to show that before the Romance 
 period the pitch accent exercised any influence whatever 
 on the sound system of Latin. On the other hand, the 
 initial stress accent has profoundly altered the original 
 vowels of medial syllables by strengthening the initial 
 syllable at their expense. Still, it is remarkable that this 
 influence is manifest only in the short vowels. Not only 
 
 
14 
 
 LATIN PHONETICS 
 
 DEVELOPMENT OF THE LATIN VOWELS 15 
 
 did the long vowels, in spite of the influence of the stress 
 accent, maintain both their acoustic quality and length, 
 but it was nothing else than the struggle between the 
 long and the stressed syllables — which gave rise to two 
 opposed rhythms — that caused the initial stress accent to 
 disappear. 
 
 {Note. — The views above given are those traditionally 
 held by French scholars. The Germans maintain, almost 
 without exception, that even in historical times the stress 
 accent was far stronger than the musical one, and they 
 explain the evidence of the Roman grammarians, which 
 is suitable only to a musical accent, as mere thoughtless 
 application of the terms used by Greek scholars when 
 discussing Greek pitch accent.] 
 
 '/ 
 
 (^ 
 
 DEVELOPMENT OF THE LATIN VOWELS. 
 
 Changes due to Initial Stress Accent. 
 
 § 8. The results of initial stress accentuation may be 
 referred to two main phenomena : — 
 
 (1) Change of the timbre or acoustic quality of short 
 
 medial vowels : umlaut, or Vowel-weakening. 
 
 (2) The disappearance or Syncope of short medial 
 
 vowels. 
 
 ^ It has not yet been possible to draw a definite line of 
 
 demarcation between these two classes : i.e.^ to explain why 
 
 nin one word the vowel has disappeared, instead of being 
 merely weakened, and vice versa. All that we can safely 
 say is that vowel-weakening invariably comes into opera- 
 tion where syncope has not taken place. 
 
 Vowel-weakening (umlaut). 
 
 § 9. The weakening of short medial vowels under the 
 
 ( I influence of the initial stress accent varied, according as 
 
 these vowels occurred in an open or close syllable, i.e.^ 
 
 ' . A according as whether only one or more consonants followed 
 
 them. We have therefore to examine their treatment 
 
 separately in either of these two positions. 
 
 A. Vowel-weakening in an Open Medial Syllable. 
 
 § 10. In an open medial syllable, except before r and 
 
 •l after i, all short vowels, without regard to their original 
 
 , acoustic quality, passed into one of the two closest short 
 
 f 
 
16 
 
 LATIN PHONETICS 
 
 V 
 
 vowels I or iX. The appearance of l or u depended on the 
 nature of the neighbouring sounds. Thus we have : — \ 
 
 (1)1 before d, t, n, ^, c; 
 
 (2) I before / followed by /, but <i before I followed 
 
 by any other vowel ; 
 
 (3) i or ti respectively before the labial sounds b, p, 
 
 /, m ; it has not as yet been possible to formulate 
 definitely the laws which decided on the choice ^ 
 of the one rather than the other. 
 
 Examples. 
 
 (1) (a) I remains : • 
 
 video ; pervldeo, evXdens, pr5vTdus. 
 
 clto (I set in motion) ; exclto, susclto. 
 
 m?nuo ; commlnuo, immlnuo. 
 
 Tlgo ; irrlguo, irrlguus. 
 
 mJco ; dimico, emico. 
 
 vicis, gen. (nom. not in use) ; invlcem. 
 (6)^>l: 
 
 sMeo ; obsldeo, praes?deo. 
 
 d^i ; reddidi, tradldi. 
 
 m^ius ; dimldius. 
 
 steti ; destiti, restlti. 
 
 stip&, nom. (from *stip?ts) ; gen. stipltis. 
 
 p^to ; compltum, * a crossway.' 
 
 t^neo ; abstlneo, sust/neo. 
 
 flum^n ; gen. flum^nis. 
 
 l?go ; sellgo, coll?go. 
 
 rego ; erigo, porr?go. 
 
 ausp^x ( = ausptfcs, from *avi -sp?c-s) ; ausplicis, gen. 
 (c) a > I : 
 
 cado ; cecidi, decldo, inc?do, succiduus. 
 
 ltd ; itldem. 
 
 fateor ; diffjfteor, proftteor. 
 
 sttttuo ; restltuo, substttuo. 
 
 datus ; edltus, prddltus. 
 
 ratus, * ratified ' ; irrltus. 
 
 efino ; cecini, concino, praecTno, tibiclnium. ^ 
 
 ttgo ; abfgo, translEgo, prodlgus. 
 
 \ 
 
 DEVELOPMENT OF THE LATIN VOWELS 17 
 
 pango (root pag) ; pepzgi. 
 tango (root tag) ; tet?gi. 
 facio ; conftcio, officio, difflcilis, aedif^cium. 
 taceo ; conticesco, ret^ceo. 
 
 facetus ; inftcetus. ^ 
 
 {d) d>l: 
 
 l^cus (in O.L. stlocus according to Quintilian, Iiistit. 
 
 Oral. i. 4, 16) ; ilrco, from *in stl5co. 
 nov^s (which became nov?/s in the Augustan period 
 
 (§ 29)) ; novltas, from *nov($tas. 
 (e) u>l: 
 
 cap?/t ; capitis. 
 
 cornil (stem corn??-) ; corn?ger. 
 
 maniis (stem man?<-) ; man?ca, 'sleeve' or * handcuff.' 
 
 (2) Exilium but ex?<lans. 
 
 fam/lia, the body of dwellers in one household, with 
 special reference to the body of slaves under one 
 master ; but famwlus. 
 
 Sicilia, but Sic??lus. 
 
 simllis, but simwlo. 
 
 Compare also — desalio, a compound of salio ; and sedttlo, 
 * with real earnestness,' from *se ddlo. 
 
 (3) habeo ; adh^beo, proh?beo. 
 taberna ; contztbernalis. 
 
 hibet ; quodllbet (from compounds of this kind a form 
 hbet established itself, which entered into competition 
 with the original form, l«tbet, and finally ousted it). 
 
 arc2*s, nom. sing. ; arcwbus, dat. abl. plural. 
 
 nian?<s ; man?bus, dat. abl. plural. 
 
 Compare also : — 
 
 intibum and intwbum, * chichory.' 
 
 auceps, nom. ; auc?ipis, gen. 
 
 manc^ps ; manci^pis, archaic form, and manclpis, classi- 
 cal form of the gen. 
 
 capio ; rectpero and recwpero. 
 
 acc^pio, but occi^po. 
 
 rapio ; surr?pio, but in Plautus surr?/pui. 
 
 aurifex and aurwfex. 
 
 manifestus (classical form) and manwfestus (archaic 
 form). 
 
 B 
 
i5 
 
 ^•^y'^f-f-r 
 
 18 LATIN PHONETICS 
 
 ferrmus, leg<mus, but possi'tmus, vol?"mus. 
 specimen, but docz^mentum. 
 
 In the case of the superlatives in -imus and -f^mus, as 
 optTmus and optr^mus, Quintilian tells us that the medial 
 vowel which was weakened represented a sound half-way 
 between t and u (Instit Orat, i. 4, 8 : medius est quidam 
 u et i litterae sonus : non enim optimum dicimus aut 
 optimum). He states also that u was the archaic and i 
 the classical spelling (/6., i. 7, 21) : jam optimus, maxtmus, 
 ut mediam litteram quae veteribus u fuerat acciperent, 
 Gai primum Caesaris inscriptione traditur factum. This 
 remark is generally confirmed by the orthography of the 
 inscriptions and the best manuscripts, which permits us 
 to apply it further to a certain number of other words 
 belonging to this third group (c/ e.g.y the genitive man- 
 cf^pis, manczpis ; maniifestus, mamfestus mentioned above). 
 At the same time it is to be noticed that as regards 
 minimus no form min2«mus appears to be proved, and 
 this fact, together with the contrast of ferlmus, legemus, 
 with possz^^mus, voir^mus, and of specimen with docwmen- 
 tum, seems to suggest a certain influence exercised by 
 the vowel of the first syllable on the acoustic quality of 
 the vowel of the second syllable. In other cases, as in 
 that of acclpio )( occwpo, it seems more probable that 
 the vowel of the syllable which followed determined the 
 acoustic quality of that which was weakened. 
 
 § 11. Before r every short medial vowel has the acoustic 
 
 value of ^. 
 
 Examples. 
 
 Norn, cinls ; gen. oin^ris (from *cinisi8, *cinlris {§ 41)). 
 
 pulvJs ; pulvft'is. 
 
 Fallsci ; FaUrii from *Fall8ii, *Fahrii (§ 41). 
 
 t 
 
 t 
 
 DEVELOPMENT OF THE LATIN VOWELS 19 
 
 legrfer ; leg^rupa, ' law breaker ' (this form is testified to 
 g by the best MSS. of Plautus, Persa, 68). 
 
 f^rus ; eff^rus. 
 verb^r ; verb^ris. 
 dare ; redd^re, tradere. 
 pario ; peperi, rep^rio. 
 Compare further— leg^re, numerus, volueris (volnus). 
 
 ^ § 12. Lastly, e represents the weakened vowel after i. 
 
 Examples. 
 
 pittas, vari^as (derived from pius and varius, older 
 forms pi5s, variSs, like novitas from novos (§ 10, 1, d)). 
 
 abietis, arietis, pari^tis, genitives from abies, aries, paries 
 (the long vowel in the nominative of these three words is 
 
 exactly comparable with that of pes ) ( pedis). 
 
 B. Vowel-iveakening in Closed Medial Syllables. 
 
 § 13. In a closed medial syllable a became ^irrespective 
 of its neighbouring sounds ; o became u from the end of 
 the third century B.C., except after w, whether vocal or 
 ^ consonantal, when it was retained in writing until the 
 commencement of the Augustan Era.^ The other vowels 
 remained unchanged. 
 
 Examples. 
 
 1. d>? : 
 arceo ; co^rceo, exerceo. 
 ^ carpo ; discerpo, exc^rpo. 
 
 scando ; asc^ndo, descendo. 
 tracto ; detr^cto, obtr^cto. 
 factus ; eff^^ctus, ref^ctus. 
 jactus ; abj^ctus, subj^ctus. 
 fallo ; fef^lli. 
 parco ; pep^rci. 
 £nnus ; bi^nnium. 
 
 ^1 
 
 1 [Probably because of the confusion entailed by the collocation uu.] 
 
 
'■?*? 
 
 ?-sp^Sv?"i«?." 'rp^^trfsr- 
 
 20 
 
 LATIN PHONETICS 
 
 ^ 
 
 
 barba ; imb^rbis. 
 aptus ; int'ptus. 
 
 O.L. end^struos, ' industrious,' properly * working within 
 one's head,' from endi5, 'within,' and striio, 'I build : c/. 
 Paulus Diaconus, p. 75, 28 Th. : industrium antiqui 
 dicebant endostruom quasi qui, quidquid ageret, intro 
 strueret et studeret domi. Classical form ind^^strlus. 
 
 mont-, stem of mons (with compensatory lengthening 
 
 of 6 (§ 25, 2)). . . 
 
 prom^nturium (this and not promSntunum is the 
 
 proper spelling). . « j 4^^ 
 
 oTl. Venos,! Veneris (the form Venos is found on two 
 
 very old mirrors) ; venr/stus. 
 
 On the other hand, during the whole Republican period 
 
 the pronunciation of the 3rd pers. plur. pros, mdic of 
 
 fruor and sequor was exclusively fruCntur, sequSntur 
 
 (§ 29). 
 
 3. I, ?, « remain unchanged : 
 dtsco ; ded/sco, perdTsco. 
 firmus ; inflrmus. 
 s^ntio ; constntio, dissentio. 
 s^vos ; cons^rvos. 
 {undo ; effrmdo, transfiindo. 
 cwrvos ; recrtrvos. 
 
 Effect of Analogy and Particular Cases. 
 
 § 14. The regular action of these laws that effected 
 the vowel weakenings was frequently crossed by analogy. 
 Thus for instance, the compounds com^^do, ded^cus, 
 posth'abeo, perfacilis, convr^co, bif.}ris, dispf^to, impMicus 
 illacrimor, subnencidus, ad()pto, etc., were remodelled 
 on the simple words .Tdo, d^^cus, habeo, facihs, yoco, foris 
 (generally in the plural foxes) pr^to, pi^^dicus, lacrimo, 
 rancidus, opto. The form which followed the natural 
 law of sound change is found sometimes retained in 
 
 1 [Properly a neuter noun.] 
 
 DEVELOPMENT OF THE LATIN VOWELS 21 
 
 popular speech. Thus, to take one example, the written 
 language recognised only the form separo (from sed, a 
 prefix denoting separation, and paro, ' I put on an equal 
 footing ' ^ (§ 57)), which was due to analogy. Certain vulgar 
 texts, however, exhibit sep^^ro, and their testimony finds 
 confirmation in the Romance languages, e.g., Fr. sevrer, 
 'to wean,' *to deprive'; separare would have produced 
 
 * sevrer [thus the Romance languages seem to postulate 
 al^Tcrem beside al«crem; cf. Grandgent, Introductwn to 
 Vulgar Latin, § 194]. 
 
 In corporis, gen. of corpus, decc^ris of decus, tempc^ris 
 of tempus, which stand for * corp^ns, * dec<?ris, "^ temp^Tris, 
 where the e should, according to rule, have been main- 
 tained, comes from the old nominatives * corpus, * decos, 
 
 * temp6»s : the regular vowel has maintained itself in the 
 adverb tempm ('at the right time,' 'punctually'). This 
 word was exempted from the influence of analogy by the 
 fact that it did not fall under the regular paradigm. 
 
 In other cases the result of the laws of vowel-weakening 
 is obscured by later changes independent of the initial 
 stress accent. In this way we must explain, e.g., attlngo 
 instead of *att^^ngo )( tango (§16); inswlsus instead of 
 ^ins^rlsus )( salsus (§ 17); seclMo instead of *seck^udo 
 )(claudo (§ 22). With regard to the last example, it 
 should be noted that of the two factors in the diphthong 
 au in claudo, the former was the more emphasised ; hence, 
 the second, w, instead of being prolonged was dropped 
 immediately, or, in other words, this u became con- 
 sonantal (§ 6, iv. 2). Thus the a of claudo stood before 
 two consonants {vd\ and the compounds of this verb 
 consequently take, according to rule, the form *-ck^udo, 
 
 1 [For pare in this Bense cf. PI. Cure, 4, 2, 20 : eodem hercle vcs pono 
 et paro : parissumi estis iibus.] 
 
22 
 
 LATIN PHONETICS 
 
 which, as already mentioned, passed, at a later period, 
 into -clildo. Compare also : — 
 
 Crti<s(s)a ; accus(s)o. 
 
 caedo (O.L. caido (§ 20)) ; cecidi : intermediate stage 
 *ceceidi (§ 19). 
 \aedo (O.L. \aido); illido : intermediate stage *ill<fido. 
 
 Syncope. 
 
 § 15. In contrast with vowel-weakening syncope does 
 not seem to have been brought about merely by initial 
 stress accent, though it is certain that this was the 
 principal factor in the process. The examples to be cited 
 immediately show indeed that the complete loss of the 
 short medial vowel was often connected with the length 
 of the following syllable. But it is not sufficient to hold 
 quantity as responsible for all the phenomena of syncope. 
 There must have been other factors as well, but in the 
 present state of our knowledge, we can say nothing 
 decisive concerning their character. 
 
 Syncope has not always entailed the loss of a syllable. 
 In fact, in those words where syncope took place in 
 immediate proximity to a i or a Vy these last sounds 
 became vowel sounds, and, unless following a vowel, 
 maintained in the word the original number of syllables. 
 
 Examples. 
 
 (a) Syncope is caused by the combined action of the 
 stress accent, and the length of the following syllable :— 
 
 ardere )( arldus. 
 
 disciplina )( discipwlus. 
 
 infra ) ( inf^rus, 'lower' {e.g., mare inferum, 'the lower,' 
 i.e., * the Tyrrhenian Sea'). 
 
 supra )( sup^rus, 'upper' {e.g. mare superum, 'the 
 upper,' i.e.y 'the Adriatic Sea'). 
 
 valde )( valldus. 
 
 ^; 
 
 DEVELOPMENT OF THE LATIN VOWELS 23 
 
 Thus the paradigm of a word like calldus was originally :— 
 
 Nom. calldus. 
 Ace. calJdum. 
 Qen. caldL 
 Dat. and Ahl. caldo. 
 
 But this original condition did not remain unchanged. 
 On the one hand analogy extended the syncopated form 
 to the nominative and accusative, and thus produced the 
 duplicate form caldus ; a form used, according to Quintilian 
 {Instit Orat, i. 6, 19), by the Emperor Augustus in 
 preference to calidus which he considered pedantic and 
 affected: ' sed Augustus quoque in epistulis ad C. Caesarem ^ 
 scriptis emendat quod is cah'dum dicere quam caldum malit, 
 non quia id non sit Latinum, sed quia sit otiosum.' On 
 the other hand analogy restored the syncopated vowel 
 in the genitive, dative, and ablative. The explanation 
 given above of caldus holds equally good for soldus, 
 * solid,' * complete,' 'the whole sum' (c/. e.g., Horace, 
 Sat i. 2, 113: 
 
 Quaerere plus prodest et inane abscindere soldo ? 
 
 and ii. 5, 65 : 
 
 Filia Nasicae metuentis reddere soldum). 
 as against soMus; raucus, 'hoarse,' from *rav(l)cus, 
 derived from r^vis, 'hoarseness,' etc. 
 
 (6) Syncope is caused by the combined action of the 
 stress accent, and of a second factor other than the 
 length of the following syUable, but which cannot be 
 defined more precisely. 
 
 abide from *abj(a)cio. 
 auceps from *av(l)ceps. 
 concutio from *conqu(a)tio. 
 
 [Son of M. Vipsaniua Agrippa, grandson of Augustus.] 
 
?i:^y'p;*»^i^ 
 
 f^ 
 
 24 LATIN PHONETICS 
 
 junior from *juv(?)nior>*juimior. 
 
 (for the contraction of *jiiunior to junior see § 32). 
 
 reccidi, reppilli, rettiili perfects of recido, repello, refero, 
 
 from *rec(?)cidi, *rep(?)puli, *ret(?)tuli (tetuli as perfect 
 
 of fero is often found in Old Latin). 
 
 In the compounds of jacio, the regular form according 
 to the law of sound change, -icio, was replaced from the 
 classical period onwards by the form -jicioy if not written, 
 at least pronounced in this way, as the metrical scansion 
 of abicioj adlcio^ conicio, obicio, subvio proves ; the 
 prefix of these words is always long by position in lyrical 
 poetry and in the classical period. Vergil, for instance, 
 begins a hexameter with obirit^ Aen., vi. 421, and vii. 
 480, and Aulus Gellius expressly remarks of subicit that 
 the i of the root syllable ' vim consonantis capit et idcirco 
 ea syllaba productius latiusque paulo pronuntiata priorem 
 syllabam brevem esse non patitur, sed reddit eam positu 
 longam,' Noct. Att., 4, 17, 8. This innovation was due 
 to analogy : coyijicio for conicio, for example, was formed 
 on a model like conficio according to the formula facio : 
 conficio =jacio : x. 
 
 Changes Independent of Initial Stress Accent. 
 
 A. Changes of the Acoustic Quality in Syllables 
 
 not Final. 
 
 § 16. ^before a guttural nasal became i. 
 
 Examples. 
 
 sept?ngenti from *sept^mcenti. 
 
 confrhigo from *confr?ngo, from *confrrtngo (§ 14). 
 
 supplngo {'I fasten underneath,' 'I beat'), from 
 *subp?ngo, from *subpango (§ 14). 
 
 d<gnus from *d?gnos, from *d?cnos (belonging to the 
 root of de'cet) : og according to § 62. 
 
 ^ j f. 
 
 K^ 
 
 «> 
 
 ^t 
 
 ^:' ' ^. 
 
 ? 
 
 
 DEVELOPMENT OF THE LATIN VOWELS 25 
 
 Ifgnum from *l?gnom (connected with l?go, ' I gather ' ; 
 lignum meant originally 'gathered wood.') [The jurist 
 Gains connected it with lego, cf. Giles, § 195, but the 
 derivation accepted by Walde is from a word akin to 
 Gk. \i.yvvs 'smoke,' ' reek.'] 
 
 For the representation of the guttural nasal by g in the 
 last two cases, see § 6 note iii., and § 69. 
 
 § 17. <9 became o before / followed by one of the vowels 
 ^, a, Oy u. In the combination -el- followed by a consonant, 
 except in the case of -//-, it passed first into o and then 
 into u (see the following paragraph). 
 
 Examples. 
 
 hSlus, O.L. helus. 
 
 The verb v^lle from *velse {§ 74), root \i\-\ vcJlo, volebam, 
 v5lam, vSlui ; but velim, v^llem, V(?lle. 
 exs?<lto from *exs^lto, from *exsalto (§ 14). 
 pulsus, perf., part. pass, from p^llo. 
 
 § 18. f>, both original o and when substituting <? (§ 17), 
 became u before / ^consonant (except in the combina- 
 tion -oil-). 
 
 Examples. 
 
 cwlmen ) ( cSlumen, a height, 
 pwlvis ) ( pSllen. 
 stwltus )( stolidus. 
 
 § 19. ei became t after the second century B.C. 
 
 Examples. 
 
 dico from dcico (deicerent is found in an inscription of 
 the year 186 B.C., the Senatus Consultum de Bacchanalibus, 
 given in the Appendix). 
 
 diffidens from diffezdens (found on an inscription dating 
 from the beginning of the second century B.C.). 
 
 incido from inceido (inceideretis in S. C. de Bacchanalibus), 
 
 c 
 
 ,'lti 
 
 -£; 
 
 ti 
 
> 
 
 26 
 
 LATIN PHONETICS 
 
 § 20. ai became ae from the beginning of the second 
 century B.C. [pronounced like ai- in * Cairo ']. 
 
 Examples. 
 
 a«des, * temple,' from aides (aide = a^dem is found on the 
 burial inscription, belonging to the third century B.C., of 
 L. Cornelius Scipio, consul 259 B.C. ; while the S. C. de 
 Bacchanalibus (186 B.C.) exhibits already a«dem). 
 
 qaacro from quazro : qua/ratis appears as an archaic 
 form beside a«tatem on the burial inscription of a Scipio, 
 dating probably from 130 B.C. 
 
 In the low country round Rome ae ceased at an early 
 period to be a diphthong. Indeed Varro informs us 
 that the peasants in his time said M^sius and ^dus for 
 Ma^sius and haedus : (de lingua Latina vii. 96 : rustici 
 pappum Mesium non Maesium ; cf. also v. 97). [An 
 attempt to change the name Caecilius to the popular 
 form Cecilius (like Claudius to Clodius), was successfully 
 laughed down by Lucilius : * Cecilius pretor ne rusticu' 
 fiat,' Lindsay, S. H. Gr„ § 2 note\ Under the Empire 
 this pronunciation gradually gained ground, and finally 
 became general. The e, arising from an older ae, was 
 open (f), in contrast with the original e, which was 
 close (f). Thus, when towards the end of the Empire 
 Latin ceased to distinguish the quantity of the vowels 
 and retained merely the differences of acoustic quality, 
 ae became completely assimilated to e {i.e., ^), while on 
 the other hand original e {i.e., ^) remained always sharply 
 distinguished from it. This is proved by the Romance 
 languages in which ae and e led to the same result, 
 whereas e was treated differently : cf. Lat. qnaerit > 
 Fr. qui^; Lat. hfTri > Fr. hier, as against Lat. c^ra > 
 Fr. ctre. 
 
 4 I f 
 
 DEVELOPMENT OF THE LATIN VOWELS 27 
 
 § 21. oi from the beginning of the second century B.C. 
 passed first into oe and then into u. 
 
 Examples. 
 
 wnus from oznos (omo = wnum in the burial inscription 
 of L. Cornelius Scipio, mentioned above) ; oenus in 
 Plautus, Truculentus, 104. 
 
 communis from commomis (comoinem in S. C, de 
 Bacchanalibus, 186 B.C.). 
 
 wsus from oiaos ; cf. Martianus Capella,^ iii. 236 : otsus 
 etiam dicitur ; sic enim veteres usum dixere. 
 
 In a certain number of words the spelling oe was 
 maintained even long after the pronunciation was u. 
 Thus mwrus is often written moerus as late as the 
 manuscripts of Vergil. Finally, by a reaction of the 
 written language on the pronunciation, oe was restored 
 in pronunciation in cases like po^a ( ) ( pwnire), moemn 
 ( ) ( mwrus), etc. 
 
 In the case of coepi the oe does not represent the Indo- 
 European diphthong oi, but is a later contraction of o + e : 
 coepi from co-^pi (for the trisyllabic scansion ^-^ cf. 
 Lucretius, iv. 619, si quis forte manu premere ac siccare 
 coepit, cf. Monro ad be), contains the perfect of a verb 
 apio, ' to join,' which seems to have become obsolete at an 
 early date [hence apiscor and aptus]. 
 
 § 22. eu passed into ou and later into u (see § 23). As 
 the transition frem eu to ou took place before the commence- 
 ment of written tradition, Latin itself shows us no example 
 of the original diphthong eu. But on the one hand, com- 
 parison with the related languages compels the conclusion 
 that the H of di^co, for example, goes back to a primitive 
 eu (the intervening grade ou is guaranteed by the form 
 
 1 [(Flor. circ. 410-427 A.D.). The African scholar who wrote the Ency- 
 Qlopedia of the Seven Liberal Arts.] 
 
 4 
 

 # 
 
 28 LATIN PHONETICS 
 
 abdowcit found in a sepulchral inscription belonging to the 
 beginning of the second century B.C. Secondly, the laws 
 of vowel-weakening entitle us to trace back acciis(s)o, for 
 example, to *adc^ws(s)o ; which in its turn arose regularly 
 from *adcaws(s)o, derived from caus(s)a, 'a lawsuit 
 
 For the apparent exceptions ceu, neu, duphcate torm oi 
 neve, s^ of sive, and n^wter, the remarks made on coepi 
 apply. In all these words eu arises from a later contraction 
 of e + w. As to nei^ter it is not even certain that this word 
 was ever pronounced in prose in any other way than as a 
 trisyllable, ne-uter. 
 
 § 23. ou passed into u towards the end of the third 
 
 century B.C. 
 
 Examples. 
 
 Incus, ' grove,' properly, ' clearing in a wood ' [c/. 
 collucare, 'to make a clearing,' connected with luceo], 
 from lawcos, which is found in an old inscription from 
 
 Spoletium. , , j j- 4.^ 
 
 N tutrix, 'nurse,' from noutrix, found in an old dedicatory 
 inscription, from Nemi [c/., however, Lindsay, .S'. H. L. (?., 
 p. 153]. 
 
 § 24. au was maintained in the written Latin of 
 Rome. But in the dialects of the adjacent districts, 
 and in the popular language of the capital itself, au 
 passed into o at an early period. This double treatment 
 of au sometimes caused confusion, as in instances 
 containing o it naturally remained doubtful if the o 
 represented a primitive o, or an au which had become 
 in popular speech. Suetonius in his biography of the 
 Emperor Vespasian, § 22, tells the following anecdote on 
 this point : (Vespasianus) Mestrium Flt^rum consularem, 
 
 # 
 
 wj 
 
 ^^ 
 
 DEVELOPMENT OF THE LATIN VOWELS 29 
 
 admonitus ab eo plawstra potius quam plostra dicenda, 
 postero die Flawrum salutavit. In doubtful cases those 
 who prided themselves on correct pronunciation, preferred 
 au, even at the risk of introducing it into words where 
 
 properly it did not belong. 
 
 In this way plawdo replaced pl(>do, although the latter 
 word contained an original d, as is shown by the com- 
 pounds complMo, displodo, explodo. In fact, if plaudo 
 contained an original au, we should have the forms 
 * compb7do, * dispb7do, * explr^do (§ 14). [au- of aufugio, 
 aufero represents an Indo-European preposition awe, 
 which was used in these compounds before an initial / 
 to avoid confusion with the compounds of ad, e.g., affero, 
 Lindsay, S. H. L. G., p. 129.-Umbrian and Faliscan had 
 in place of Latin au : so, in general, the dialects of 
 T^orthern and Central Italy. Grandgent, § 211.] 
 
 B. Changes of Quantity in Syllables not Final. 
 (a) Lengthening of Short Vowels. 
 § 25. A short vowel received compensatory lengthening— 
 
 (1) In cases of the loss of an s before voiced consonants 
 
 (§ 86). 
 
 Examples, 
 
 dinosco from *dIsnosco. 
 t'duco from *exduco {x = cs). 
 idem from ♦is-dem (is + dem). 
 
 pono from *pSs(i)no ; cf. the participle positus = po 
 (prefix) + situs, perf. part. pass, of sino. 
 sedecim from *sexdecim. 
 
 (2) Before -nf- and -ns-, because in this combination 
 the nasal lost its character as stop, and the vibrations 
 
30 
 
 LATIN PHONETICS 
 
 of the ligaments of the glottis that caused the sound 
 were added to the preceding vowel (§ 87). 
 
 Cicero, Orator , 48, 159 : quid vero hoc elegantius, quod 
 non fit natura, sed quodam instituto, mdoctus dicimus 
 brevi prima littera, Znsanus producta, mhumanus brevi, 
 mfelix longa et, ne multis, quibus in verbis eae primae 
 litterae sunt quae in * sapiente ' atque ' felice,' producte 
 dicitur m, in ceteris omnibus breviter : itemque c^nposuit, 
 cJnsuevit, concrepuit, ccJnfecit : consule veritatem, repre- 
 hendet; refer ad aures, probabunt. These remarks of 
 Cicero are confirmed by the inscriptions which show at 
 the same time that the lengthening process in question 
 was not confined to the i of the prefix m- and the o of 
 the prefix con-. In fact, in the Latin epigraphic records, 
 every kind of vowel occurring before one of the two com- 
 binations -nf' and -ns- is frequently marked as long by a 
 special sign, the so-called apex. Further, as the Greek 
 alphabet possessed two different letters for e and e and 
 also for and J, the Greek transcription of Latin words 
 in inscriptions and in authors may also to some extent 
 give us information as to the quantity of vowels before 
 -nf' or -ns; and this evidence is absolutely in favour of 
 their having been long. 
 
 Amongst other instances which might be cited, a passage 
 in Quintilian {Instit. Orat.^ i. 7, 29) proves that the fore- 
 going are really cases of compensatory lengthening. He 
 remarks that in the word cdnsides the nasal n was not 
 pronounced (. . . consules exempta n littera legimus) [just 
 as it disappeared in 0. Eng. fif (Germ, funf), gos, ' goose ' 
 (Germ, gaws), Lindsay, p. 11]. 
 
 § 26. Irrespective of the lengthening by compensation, 
 a short vowel was lengthened in perfect participles passive, 
 
 h 
 
 ^) 
 
 4. 
 
 (0 
 
 . II « 
 
 DEVELOPMENT OF THE LATIN VOWELS 31 
 
 in the case of verbs whose root ended with a voiced 
 consonant, and again in the verbal and noun forms derived 
 from these participles (§62, 1). 
 
 Examples. 
 
 actus, lectus, part, from ago, l^go. 
 
 (The length of the root syllables in these two participles 
 is proved by the testimony of Aulus Gellius, Koctes 
 Atticae, ix. 6, and xii. 3, and also by the apex in numerous 
 inscriptions). 
 
 casus, visus, participles from cado, video (§ 83), but 
 
 fact us (participle of facio). 
 
 (If the a of factus had been long, the compounds would 
 have been *confactus,-2effactus, etc., instead of cdnfectus 
 effectus, as is the case, cf. adactus.) 
 
 ra^sus, part, of meto (§ 83). 
 
 (That the ? in mcssus was short may be shown as 
 follows: towards the end of the first century B.C., -ss- 
 following a long vowel was simplified into -s- ; conse- 
 quently had messus contained a long e, it would have at 
 that period appeared in the form *mesus, which is not 
 the case (§ 58).) 
 
 In the same way we have fictito, frequentative of ago 
 [cf. ambages], protector, as against factito, s^tio. 
 
 Exceptions like s^ssus, part, of s^eo, fSssus, part, of 
 f^dio, arose probably under the influence of analogy. 
 
 (b.) Shortening of Long Vowels. 
 
 § 27. In the classical period every long vowel that was 
 followed, but not preceded by another vowel, was shortened. 
 
 Examples. 
 
 f inlo ) ( the infinitive f inire. 
 
 pius (in O.L. pius : a hexameter of Ennius begins with 
 the words pectora pia tenet desiderium). 
 
 r^ gen. of res (but diei gen. of dies, glaciei gen. of glacies, 
 because in these words the e was not only followed but also 
 preceded by a vowel). 
 
 

 VtjfTBSF??-"- 
 
 32 LATIN PHONETICS 
 
 ago ) ( the imperfect flcbam. 
 
 gruis, gen. of grfis, ' a crane.' 
 
 s«o )( sstor, ' a cobbler ' [sSbula, ' an awl ]. 
 
 The classical poets scan as a rule iUms, ipsTus, unlus ; 
 but in prose the accepted pronunciation at least in 
 Quintilian's time, was ill7us, ip87us, un7us (InsM Orat., 
 i 5 18 unTus extra carmen non deprendas). Ihe 
 reasons for the restoration of the long quantity in 
 these genitives have not been sufficiently elucidated. 
 [It has been suggested that they may be due to the 
 long vowel I in the datives ill?, ips7, un7.] On the other 
 hand when Servius, the commentator on vTergil, informs 
 us that in the perfect of audire and lenire prose writers 
 insisted on aud7it, len7it, in opposition to the scansion 
 audlit, lenTit of the poets, the infli^nce '^^^^^'^jyj' 
 obvious : aud7it and len7it are formed after their dupli- 
 cate forms audZvit, len7vit, in which cases the 7 was not 
 followed by a vowel, and hence could not be shortened. 
 In the same way f7o is a case of substitution for ho, 
 
 following f 7s, f 7mus. . „ , i • 
 
 (For a difficult case of the shortening of a long vowel in 
 a syllable not concluding a word see below, § 55). 
 
 a Changes of Acoustic Quality in Final Syllables. 
 
 8 28 In cases where a absolutely terminates a word it 
 remains unchanged; compare genera, frigor«, nom. andacc. 
 plur. of genus and frigus; it«; where the short quantity 
 of final « is proved by the evidence of the other Indo- 
 European languages to be original. In other cases final a 
 is secondary, i.e., it comes from the shortening of an « 
 (see puta below, §31, 3). , , i. 
 
 With re-ard to the other original short vowels when 
 absolutely final, they seem one and aU to have taken on 
 
 ^ 
 
 i' 
 
 I V 
 
 
 DEVELOPMENT OF THE LATIN VOWELS 33 
 
 the acoustic quality of e. Linguistic testimony, however, 
 is too scanty to provide positive proof except in the case 
 of «, cf. mar^, nom. and ace. sing. ) ( mar?a nom. and ace. 
 plur. ; len<9, neuter ) ( lems, masc. and fern. 
 
 § 29. When not absolutely final, the short vowels were 
 treated quite differently. In the case of words of one 
 syllable they underwent no change. In words of two or 
 more syllables ^ and ii remained, and likewise e, except 
 before s, when it became ^; d before two consonants 
 developed in the same way as in a medial syllable, 
 i.e., it passed into e; as to its treatment before a single 
 consonant, nothing certain can be said in the absence of 
 decisive examples. Finally, o became u in the nom. and 
 ace. sing, of the 2nd declension, in the nom. and ace. 
 sing, of neuter substantives of the 3rd declension, and in 
 the 3rd pers. plur. indie, of the present and perfect of 
 verbs. This change of o goes back to the end of the 
 third century B.C., in words where o was not preceded by 
 u or V. After u and v, o changed to u only after the 
 beginning of the Empire. 
 
 Examples : Monosyllables. 
 
 n?x, n?vis ; pfx, pi^cis ; m^l, mollis ; n^x, n^cis ; lac, lactis 
 fax, facis ; ^s, ^ssis ; nCx, n^otis ; nwx, nwcis ; tr«x, 
 triicis. 
 
 Dissyllabic and Polysyllabic Words. 
 
 {a) (1) ensis (stem ens?- ; cf. gen. plur. ens?-um) ; calix, 
 
 calicis ; agiUs (stem agiU-). 
 (2) turtwr, -i^ris, turtle-dove ; magistratus, -us (stem 
 
 magistratw-) ; redwx, redwcis. 
 (6) flum^^n, -Inis ; forf^x, -?cis ; hi?ms, hiftuis ; harusp^x, 
 
 -?cis (specie in O.L.); hiU generis from* generis, gen. 
 
 of genus ; ordinis from *ordinfe, gen. of ordo ; just 
 
 as in the genitives attested by old inscriptions 
 
 Apoloncs = Apollin?s, Cereres = Cererls, Salutes = 
 
 SalutiS, Veneres = Veneres. 
 
 
 1 
 
 "%, 
 
34 
 
 LATIN PHONETICS 
 
 Cases like des^s, -sTdis, div.^s, -Itis are merely 
 apparent exceptions. In fact des^ ^^^^J'^f.^^ 
 a form MesM-s, and div^ to a form Miv?t-s (§ 68). 
 Thus the ? of words of this type did not, as m the 
 examples just given, stand originally before s ; and 
 this fact explains its retention, 
 (c) auc^ps from *av(i) -cap-s (avis + capio). 
 rerafx from *rem-rtg-s (remus + ago), 
 aurif^^x from *auri-frtc-s (aurum + facie). 
 {d) fili?/s, from filieJs, nom. sing. 
 \ivum from virom, ace. sing, of vir. 
 donilm from donSm, nom. and ace. sing. 
 opus from op3s, nom. and ace. sing, 
 consentient from consentient, 3rd pers. plur., present 
 
 indie, consentio. 
 dederrmt from dederflnt, 3rd pers. plur., perf. mdic. of do. 
 The forms fili(5s, virom, donom, op5s, co(n)senti5nt, 
 deder^Jnt, are testified to partly by inscriptions partly by 
 quotations in Latin authors, c/., for example, Qumtilian, 
 InstiL OraL, i. 4, 16. With regard to words like biduSm, 
 equ^s viv^nt, modern editions of Caesar, for example, 
 ought not to give the forms bidu^an, equws, vivimt. In- 
 scriptions prove unmistakably that these forms with u 
 were later than the Republican era ; thus their introduc- 
 tion into a text like the Bell. Gall, is an anachronism.^ 
 
 § 30. Long vowels, whether absolutely final or in the last 
 syllable, suffered no qualitative changes. On the other 
 hand, the diphthongs which occurred in either of these 
 positions became long vowels. Of these it is especially 
 important to note (i.) that ei, ai, oi became 7 ; and (u.) 
 that ou became u. 
 
 Examples. 
 
 sib?, dat. of the reflexive pron. from sihei. 
 
 abis, 2nd pers. sing. pres. indie, of abeo from *abas. 
 
 tutudi, perf. of tundo, from *tutudat. 
 
 I 
 
 «/ 
 
 >ll 
 
 DEVELOPMENT OF THE LATIN VOWELS 35 
 
 ros2S, dat. and abl. plur. of rosa, from *rosazs. 
 
 • viri, nom. and voc. plur. of vir, from *viroi. 
 
 donJs, dat. and abl. plur. of donum, from *dono/s. 
 
 currws, gen. sing, of currus, from *curro2/s. 
 
 Sometimes the original diphthong is attested by inscrip- 
 tions {e.g. sihei appears in the S. C. de Bacch.), or by 
 quotations made by grammarians (as by Paulus Diaconus,^ 
 p. 14, 17 Th. : ab oloes dicebant pro ab illis), but generally 
 
 ,^ speaking its existence is revealed only by a comparison 
 
 with the other Indo-Euroj^ean languages. 
 
 As then at was represented by i, the termination ae in 
 the gen. and dat. sing, and in the nom. and voc. plur. 
 of words of the 1st declension cannot have arisen from ai. 
 As a matter of fact it represented rather an older at ; cf. 
 ,|^ the archaic gen. aulai in Vergil, Aen.^ iii. 354, auki' medio 
 libabant pocula Baccho, and again pictcTZ, Ae?h, ix. 26 ; 
 auraJ, vi. 747, aqu«7, vii. 464, all gen. sing, [so IphianassaJ 
 in Lucret., i. 84, 6 ; Martial ridicules the form ai as 
 affected, xi., 90, 5 ; attonitusque legis ' terrai frugiferai,' 
 see S. H. L. G., p. 44]. 
 
 £>. Changes of Quantity in Final Syllables. 
 
 § 31. The vowel element of a final syllable tended to 
 undergo abbreviation of length. This is shown in the 
 following cases : — 
 
 1. Every short vowel standing absolutely last was 
 
 ♦ reduced to a fraction of the unity of length, and accord- 
 ingly showed a tendency to disappear entirely. 
 
 Examples. 
 
 The imperatives die, due, fac, fer )( cap^^, mitt^, tund^, etc. 
 
 The duplicate forms ac (§ 82) and atqu^; nee (§ 82) and 
 
 neque (it should be noted that ac and nee were generally 
 
 h 
 
 1 [But the pronunciation from the third century B.C. was probably u 
 not 0, cf. Velius Longus, vii. p. 48, 4k, Hermann.] 
 
 1 [Paulus Diaconus, 725-797 A.D., a Benedictine monk who wrote a 
 summary of Festus' (second century A.D.) abridgment of Verrius Flaccus 
 (arc. 10 B.C.) De Verborum Signijicatu.] 
 
'iSUtV^ .(Tj^jiv. ■ 
 
 t 
 
 36 LATIN PHONETICS 
 
 used before a word that began with a consonant, and that 
 hence the disappearance of terminal ^ cannot be due to 
 elision) ; neu and nev^. 
 tot from *totl'; cf. totidem. 
 
 2. Every long vowel was shortened before any consonant 
 other than s. 
 
 Examples. 
 
 animal, gen. animalis ; calcar, gen. calcaris. 
 lictyr, gen. lictoris. 
 
 cantabam, but cantabas ; punit, but punis ; spem, ace. of 
 spes. 
 
 3. In cases of absolute finals of iambic dissyllables 
 Old Latin poetry could count every long vowel as short ; 
 and there is no doubt that this rule of scansion reflected 
 the ordinary pronunciation of the age. On the other 
 hand, the classical usage admitted the scansion ^^ only 
 in the case of certain words used to help out a phrase, 
 and so constantly employed that they were actually 
 pronounced less completely than the other words in the 
 phrase. 
 
 Examples. 
 
 ben^, mal^)( cate, slyly, fere. 
 
 cit6, mod6){ eo, * thither,' retro. 
 
 puta, 'for instance,' properly 'bring into a line of 
 reckoning,' imp. of putare; hav?, 'hail!', properly 
 'rejoice,' imp. from (h)av^re (the pronunciation ave was 
 characterised by Quintilian as pedantic, Instit. OraL, 
 i. 6, 21), but ama, doce. 
 
 quas?, adv. and conj. from quast (quasei is found on 
 several inscriptions of the second century B.C. ; ei > i 
 
 (§ 19). 
 [Cf. also such scansions as cav^-fdcias, vol^-scire, as well 
 
 as volwptatem, eg^statem, fert^ntarium, where the syllable 
 
 scanned as a short syllable is long, not by nature, but 
 
 by ' position ' , this shortening process which turns an 
 
 A: 
 
 DEVELOPMENT OF THE LATIN VOWELS 37 
 
 iambus into a pyrrhic in the Latin dramatists is called 
 the law of the Brevis Brevians. An example is found 
 in Catullus (Ixi. 77)— 
 
 virgo adest, vid^n ut faces 
 splendidas quatiunt comas.] 
 
 E. Contraction of Voivels. 
 
 § 32. When two adjacent vowels made a hiatus, that 
 is to say, when the change from the first to the second 
 was brought about by a sharp interrupted movement in 
 the utterance, they were as a rule contracted, i.e., they 
 coalesced into a single vowel or a diphthong. That was 
 the case in combinations of the type i + i, e-\-e, a + a, o-ho, 
 u + u, + ?, o+e, etc. When, on the other hand, the 
 transition was effected by a light gliding movement with- 
 out break of continuity — in other words, when there was 
 no hiatus — the two vowels remained uncontracted. This 
 was the treatment of the combinations i+ani/ other vowel 
 except iy and u + aiiy other vowel except u. However, the 
 combinations e-{-a and e + 6, although belonging to the 
 former of the two categories just mentioned, seem never 
 to have undergone contraction. 
 
 The most obvious and common examples of this process 
 are the following : — 
 
 nil from n?{h)tl ; m? from m?(h)?. 
 
 eH-e=e. 
 
 nemo from * n^-(h)emo (hemo = homo is attested by Paul us 
 Diaconus, p. 71, 18 Th.). 
 
 demo from * de-^mo ; degi, perfect of dego (contracted from 
 * de-ago (see below) from * de-egi). 
 
 a-\-d = d. 
 
 latrina from la(v)atrina. 
 
 copia from * ci^-dpia {cf, in-6pia). 
 
• I * 
 
 38 LATIN PHONETICS 
 
 coram, adv. and from Cicero's time preposition, from * c3- 
 oram (the second part of the compound is connected 
 with OS, oris; for the termination cf. clam and palam), 
 [These terminations are adverbial ace. sing. fern, forms,] 
 
 ii + iizz u, 
 
 jwnior, comparative from juvenis, from *]uumoT (§ 15&). 
 
 e + a = €. 
 
 dego from * de-ago. 
 
 5 + i = oe. 
 
 coetus, a gathering, from c^-Itus. 
 
 como from * co-emo ; contio from * co(v)?ntio (§ 50). 
 
 ij + e = oe. 
 
 coepi from c^-epi (§21). 
 
 6 + d = 6. 
 
 cogo from * co-ago. 
 
 copula, from * co-«i)ula (apere in O.L. meant 'to fasten' 
 (§ 21). 
 
 In a number of cases analogy restored the uncontracted 
 form. Thus \m\\i was contracted according to rule to m?, 
 but restored later on the analogy of tibi and sibi which 
 always remained disyllabic. 
 
 The same remark applies to d^i^sse beside d^sse, diedmo 
 ) ( d^go, cortlesco ) ( C(>go, etc. The contrast between the 
 present cogo and the perfect Qoegi, the latter of which 
 was restored by analogy, while the former escaped this 
 influence, may be due to the fact that in cogo the con- 
 traction goes back to a very old date, while the coalescence 
 oi d-\-e into oe seems, on the contrary, to be comparatively 
 recent ; thus the perfect was more accessible to the influ- 
 ence of analogy than the present. It should be added, 
 however, that generally speaking the chronology of the 
 laws of Latin contraction is an exceedingly obscure subject, 
 and it seems improbable that it will ever be satisfactorily 
 elucidated. 
 
 Synizesis must not be confused with contraction. By 
 
 • f ♦ 
 
 e 
 
 e 
 
 DEVELOPMENT OF THE LATIN VOWELS 39 
 
 synizesis is meant the liberty taken in prosody, according 
 to which two vowels in juxtaposition, although both pro- 
 nounced, count in scansion as a single syllable : c/. Vergil, 
 Geologies J iv. 34 : 
 
 Seu lento fuerint alve'aria vimine texta. 
 
 Aen. vii. 190 : 
 
 Aure'k percussum virga versumque venenis. 
 
 Ovid, Met, ix. 143. 
 diffudit miseranda suom ; mox deinde quid autem. 
 
 Vergil, Aen.y i. 131 : 
 
 Eurum ad se Zephyr unique vocat, deBmc talia fatur ; 
 but compare, e.g , 
 
 Vergil, Georg,^ iii. 167. 
 cervici subnecte j dchinc, ubi libera colla. 
 
 Ablaut or Vowel Gradation. 
 
 § 33. Besides the vowel weakening apparent in conHcio ) ( 
 facio ; exswlto ) ( salto ; volo ) ( v^lim ; infra ) ( inf^rus ; r^m 
 ) ( r^s, etc., which, as we have already seen, have their 
 origin in the separate development of Latin, and conse- 
 quently are peculiar to this language, there are other 
 vowel changes which are found also in the allied languages, 
 and presumably go back to the Indo-European. As a 
 matter of fact, from the period when the Indo-European 
 stock was still undivided, each of the three component 
 parts of a word, i.e.^ root, suffix, and termination (pp. v, vii) 
 could assume diff'erent types of vowel system to indicate 
 different types of formation. This morphological process, 
 originally closely adhered to, is known under the name of 
 ablaut or vowel gradation. 
 
 Ablaut is strongly marked in Greek ;^ in Latin its 
 influence has been largely obscured. As a consequence of 
 
 1 [As also in German and, to a less extent, in English,] 
 
h'^^- --"^^ 
 
 
 40 
 
 LATIN PHONETICS 
 
 circumstances which we cannot examine here, the original 
 conditions obtaining in this particular branch of the Indo- 
 European family of languages have been so greatly changed 
 that the system of its vowel gradations has become un- 
 recognisable. We must therefore content ourselves with 
 the mere enumeration of a few of the most characteristic 
 
 instances. 
 
 Examples. 
 
 (i.) The P^ grade combined with the o grade. 
 
 p^ndo, ' I weigh ' ; pSndo, indecl, ' in weight,' abl. of an 
 obsolete substantive * pondus -i. 
 t^go ; tSga. 
 equ?, voc. ; equcJs, nom. 
 
 (ii.) The ^ grade combined w^th the nought grade. 
 ?do, I eat ; d- ens, tooth, orig. pres. part, of Mo. 
 est ; s -unt. 
 ggn- ui ; gi- gn— o. 
 (iii.) ^ grade 6- grade and nought grade combined. 
 
 O.L. f?ido (Class. Lat. fzdo (§ 19); O.L. fSidos (the abl. 
 foidere is found in an old inscription. Class. Lat. foedus) ; 
 fides, 
 (iv. ) The e grade combined with the P grade, 
 emi perf. ; t^mo pres. 
 tegula; t^go. 
 (v.) The ? grade, e grade, 6 grade and nought grade are com- 
 bined in sMeo sedes solium (for d > I see § 37) sido from 
 *si-sd-o (a formation like gi-gn-o above): *sisdo > 
 *sizdo > sido according to §§ 25 and 86. 
 
 (vi.) The grade combined with the ? grade. 
 
 O.L. raajosem (Class. Latin majorem) ; maj^stas (the S 
 grade is concealed in the neuter raajus, older *maj5s). 
 O.L. *honosem. Class. Lat. honorem ; hon^stus. 
 
 (vii. ) The o grade combined with the 6 grade. 
 fodi perf. ; ffidio pres. 
 odi ; 3dium. 
 
 1 [The e and a grades are known as High, the o grade as Low, the nought 
 grade as Weak. The last was due to stress accent.] 
 
 f. 
 
 4r. ^ # 
 
 >; 
 
 '^ 
 
 ^j 
 
 1% ^ % 
 
 DEVELOPMENT OF THE LATIN VOWELS 41 
 
 (viii. ) The o grade combined with the nought grade. 
 
 ne- potem, ace, sing, of nepos (' grandson ' or ' nephew' ) ; 
 ne-pt-em. ace. sing, of neptis (' granddaughter ' or * niece '). 
 genitorem ; gene-tr-icem. 
 (ix. ) The e grade combined with the a grade, 
 feci ; facio. 
 semen; satus. 
 (x. ) The grade combined with the a grade. 
 
 cos, a hone ; catus, properly ' sharpened,' and so ' sly.' 
 donum; datus. 
 
 1.^ 
 
 Lj'' 
 
 < L 
 
■e-TfTrys,-' 
 
 ??^*S^J^?^i^?7E5fS^ 
 
 42 
 
 LATIN PHONETICS 
 
 HISTORY OF THE LATIN CONSONANTS. 
 
 Single Consonants. 
 
 Stops. 
 
 § 34. When between two vowels, the Latin stops, more 
 especially the breathed stops, underwent but few changes. 
 The following are worthy of remark : — 
 
 Breathed Stops. 
 
 § 35. From the end of the second century B.C. the fashion 
 became prevalent of introducing into a number of genuine 
 Latin words the aspiration of the three kinds of breathed 
 stops. Before that date this peculiarity was confined to loan- 
 words from Greek (§ 6, note iv. 4). When Greek civilisation 
 was transplanted to Rome, and attracted a constantly- 
 increasing number of adherents, certain Roman families 
 thought it elegant to give their names a Greek stamp by 
 adorning them with aspiration. Thus Sempronius altered his 
 cognomen Graccus, ' a jay,' into GraccAus, to rhyme with 
 Bacchus: we find on coins of the year 103 B.C. PulcAer; 
 and again the names Ce^egus, O^o, Tbrius became Ce^Aegus, 
 Otko, Ti^orius. From proper names this habit spread to 
 common nouns. Thus instead of ancora, lacrima, pulcer, 
 sepulcrum, sul^>>ur, etc., it became the fashion to say 
 ancAora, lacArima, pulcAer, sepulcArum, suljt?Aur, etc. We 
 possess a body of instructive evidence on this practice in 
 passages from ancient authors, of which we may cite the 
 two most important: Cicero, Orator, 48, 160; quin ego 
 ipse, cum scirem ita majores locutos, ut nusquam nisi in 
 
 (■ 
 
 
 I 
 
 HISTORY OF THE LATIN CONSONANTS 43 
 
 vocali aspiratione uterentur, loquebar sic, ut pulcros, 
 Ce^egos, triumjt?os, Kar^aginem dicerem ; aliquando, idque 
 sero, convicio aurium cum extorta mihi Veritas esset, usum 
 loquendi populo concessi, scientiam mihi reservavi. 
 Orcivios tamen et Ma^ones, Ozones, Caepiones, sepulcra, 
 coronas, lacrimas dicimus quia per aurium judicium licet : 
 Quintilian, Instit. Orat., i. 5, 20; diu deinde servatum, 
 ne consonantibus [veteres] adspirarent, ut in Graccis et in 
 triumj[?is, erupit brevi tempore nimius usus, ut cAoronae, 
 c/ienturiones, praecAones adhuc quibusdam in inscrip- 
 tionibus maneant, qua de re Catulli nobile epigramma 
 est. The epigram of Catullus referred to by Quintilian is 
 No. 84, which begins with the words ; — 
 
 C/iommoda dicebat si quando commoda vellet 
 dicere, et insidias Arrius /dnsidias, 
 
 [Q/*. also Quintilian, xii. 10, 27, and Lindsay, S. H, L. G., 
 p. 16: ' Nigidius, a grammarian of Cicero's time, empha- 
 sised the importance of accuracy in the use of the letter h : 
 rusticus fit sermo si adspires perperam ; St Augustine says 
 the dropping of h was a sign of vulgarity ; Gellius says 
 a^enum, veAemens, incoAare are archaic, see Grandgent, 
 Introduction to Vulgar Latin, p. 106.] 
 
 No trace of this affectation remains in the Romance 
 languages ; thus we may probably conclude that the 
 fashion, after degenerating into a mania, eventually fell 
 into complete disuse. 
 
 It is important to note that the sound of Latin cA, ph, 
 and th was the same as that of German k, p, t before 
 vowels; [i.e., the aspiration was similar to, but not so 
 strong as, that heard in par^-Aead, iop-hdX, foo^Aold.] 
 We must therefore reject the pronounciations /Itrum, 
 lim/a, sul/ur. 
 
 
 »*J^Si^f£i^ 
 
 y^ 
 
f> 
 
 ■:7~T<r--rT' 
 
 44 
 
 LATIN PHONETICS 
 
 Voiced Stops. 
 
 b. 
 
 § 36. b became a spirant from the second half of the 
 first century a.d., and was pronounced like the N. German 
 w in words like Wein^ or the French v in words like vin. 
 This change first set in when b was in the middle of a 
 word between two vowels, but afterwards also when b was 
 initial, though only in one part of the Roman Empire. 
 Hence on inscriptions we find constant interchanges of b 
 and V (i.e., consonantal w), which also about the same period 
 had become a spirant (§ 52). Thus we find, for instance, 
 incomparavilis for incompara^ilis ; libertavus for liberta6us 
 — from liberta ; rene for ^ene, and, on the other hand, ^eni 
 for veni, bixi for rixi, le^are for lerare. This confusion 
 increased as time went on, so that in the fifth century a.d. 
 a grammarian found it necessary to write a special treatise 
 entitled ' De b vocali et v vocali.* This change is reflected, 
 too, in the Romance languages, cf. Fr. ^esan^on, from Lat. 
 Fesuntionem ; devoir from de^ere ; fSre from fa6a.^ 
 
 d. 
 
 § 37. d was replaced by / in O.L. c?acrima> 
 Class. Lat. /acrima (cf. Paulus Diaconus, p. 48, 15 Th. : 
 (/acrimas pro /acrimas Livius [Andronicus] saepe posuit), 
 O.L. c^autia, -iorum, hospitality offered to foreign ambas- 
 sadors > Class. Lat. /autia. Paulus Diaconus says that 
 Livius Andronicus often used this word for lautia, p. 48, 
 16 Th. ; lautia is frequently used by liivy, cf. e.g. xxx. 
 17, 14. O.L. c?ingua>/ingua (Marius Victorinus,^ vi. 
 
 1 [See Grandgent, p. 134. 6 is substituted for v much oftener than v for 
 b ; other common instancei are salium, sertus, solvit : App. Prob. * alveus 
 
 non albeus. '1 ... 
 
 2 [Marius Victorinus, a grammarian and rhetorician of African origin, 
 floruit circ. the middle of the fourth century a. D.J 
 
 i 
 
 0i> 
 
 rr 
 
 V 
 
 HISTORY OF THE LATIN CONSONANTS 45 
 
 p. 9, 17k, says : nos nunc . . . /inguam per ^potius quam per 
 d [scribamus]). In some of these cases it is supposed that 
 analogy has been operative, as in the case of o?inguavpossibly 
 influenced by lingo, 'I lick'; /autia by /autus, '^^fty'; 
 it has also been thought that the substitution of / for d 
 was due to Sabine influence. 
 
 Sometimes d and / change in forms which belong to the 
 same root ; e.g., odor, o/ere [perhaps influenced by o/eum], 
 sec^eo, so/ium (for the vowel gradation, see § 33). 
 
 § 38. Final d remained after a short vowel ; but after a 
 long vowel it disappeared from the beginning of the second 
 century B.C. 
 
 Examples. 
 
 1. ac^prep., aliuc?, neuter of alius; illuo?, neuter of ille ; 
 sec?, conjunction. 
 
 2. The ablative singular of all five declensions [and of 
 ego, tu, sej; the 2nd pers. sing, and 3rd pers. sing, and 
 plur. of the imperative of verbs. 
 
 (a) praeda, abl. sing, of praeda, O.L. praidarf. 
 naerito, earlier nieritorf. 
 
 aeri (replaced at an early date by aere, which, like 
 all ablatives in -^ of the third declension is really 
 an old instrumental case), abl. sing, of aes, 
 O.L. airirf. 
 
 magistrate, abl. sing, of magistratus, O.L. magis- 
 tratuc?. 
 
 die, abl. sing, of dies, O.L. *diec?. 
 
 Praidarf, meritoo?, airic?, magistratuc? are attested 
 by old inscriptions ; for the fifth declension we 
 have no evidence. C/., too, Naevius, Bellum 
 Punicum, 7 m. : — 
 
 Noctu Troiac? exibant capitibus opertis. 
 
 [med, ted, serf, old forms of abl. sing. ; for long 
 vowel, see Giles, § 328, iv.] 
 
 (b) dato, 2nd and 3rd pers. sing, of the imperative of 
 
 dare, from original datorf. 
 
 I 
 
 itSitedibiiaaagiiiiiiaaiBatiftMia«Miia 
 
i 
 
 46 LATIN PHONETICS 
 
 sunto, 3rd pers. plur. imp. from esse, from original 
 suntorf. [This -tod seems originally to have been 
 the abl. sing, of the demonstrative pronoun -to- 
 * from that,' ' thereon,' and hence abito will 
 mean *go away at once,' or 'be sure you go 
 away,' a stronger form of command than abi 
 (see Lindsay, S.H.O., p. 103; Giles, § 519).] 
 datod and suntod are attested by inscriptions. 
 
 The negative hauc? maintained its c?, although the 
 preceding diphthong au was in the duration of its pronun- 
 ciation equivalent to a long vowel, because the word was 
 proclitic ; i.e., attracted into the following word, so that its 
 d was not final. 
 
 § 39. The pure voiced guttural g suffered no changes. 
 On the other hand, the labialised voiced guttural which 
 we represent by g^\ and which the Latin orthography 
 represented by gu^ was replaced by v whenever it occurred 
 between two vowels. 
 
 Examples. 
 
 niris, from nix, which goes back to an original form, 
 *nig's (§ 82), )( nin^wit, 'it snows.' 
 
 struyo, written struo (§ 47), )( structum, the supine, 
 which developed from *stru^'' tum (§ 82). 
 
 It is important to note that g^ and q^, written in 
 ordinary spelling gu and qu, must not be regarded as 
 consonantal combinations. They were not groups formed 
 of a guttural stop and the semi-vowel v, but guttural 
 stops with a labial aftersound ; the latter receiving a very 
 much weaker articulation than the semi-vowel v. 
 
 Thus qu standing alone is not sufficient to make a 
 syllable long by 'position' (§ 95), as would be the case 
 if the combination were regarded as one of two independent 
 
 f\ 
 
 y' 
 
 c 
 
 '*) 
 
 HISTORY OF THE LATIN CONSONANTS 47 
 
 consonants: cf. Vergil, Ae?2., ii. 15, Instar mentis ^quom 
 divina Palladis arte. 
 
 Spirants. 
 
 / 
 
 § 40. /was not tolerated in the middle of words ; for the 
 medial / of the other Italian dialects, Latin substituted 
 b or d. Still there are a few cases in which / is actually 
 found in this position, cf. e.g., fe/elli, de/ero, re/icio, ru/us, 
 scro/a, va/er. 
 
 The explanation of these exceptions is simple. Cases 
 like those of fefelli, defero, reficio are to be explained by 
 analogical restoration, and rufus, scrofa, vafer are loan- 
 words from neighbouring dialects adopted into the language 
 of Rome. We find such loanwords frequently in the 
 language of the people, where the literary language main- 
 tained the genuine Latin form. Thus we gather from a 
 passage in Nonius Marcellus,i (p. 531, 2 m.), that the 
 educated Romans pronounced si^ilare, while the people 
 said si/ilare : si/ilare quod nos, vilitatem verbi vitantes, 
 siiilare dicimus. From the form si/ilare came the 
 French word ' siffler.' In the same way the Latin writers 
 call the truffle tuber, while the Romance languages, as 
 the French word triiffe proves, imply the existence of a 
 duplicate vulgar form ^tu/er. 
 
 § 41. The breathed dental spirant s, when beginning a 
 word, remained unchanged. On the other hand, when 
 occurring between two vowels in the middle of a word, it 
 
 1 [Nonius Marcellus, a native of Numidia, floruit circ. 323 a.d. He is 
 the author of an encyclopaedic work called De compendiosa doctrina. 
 See Sandy's History/ of Classical Scholarship, vol. i. p. 209.] 
 
 S2 
 
 |i 
 
48 
 
 LATIN PHONETICS 
 
 first became voiced, and then passed into r. Instances of 
 this process, commonly known as rhotacism, are very 
 numerous. Cf. aens, gen. of ae^*?; juris, gen. of ju.?; opens, 
 gen. of opu5 ; feriae, 'holidays,' ) ( fe^tus ; heri ) ( hesternus; 
 haurio, ) ( supine hau^tum ; queror, ) ( questus sum ; erit, 
 future, ) ( e5t, present ; dirimo, older form Misemo ; amare, 
 delere, leg^re, audi re, ) ( es^e. 
 
 We may compare Varro, de lingua Latina vii. 26 : in 
 multis verbis, in quo antiqui dicebant 5, postea dicunt r . . . 
 foede^um, foederum ; plu^ima, plurima; meliosem, melio7'em ; 
 a^enam, arenam ; Quintilian, Instit. Orat. i. 4, 13 : nam ut 
 Vale^ii Fu^ii in Valerios Fu?'iosque venerunt, ita . . lases 
 et asa fuerunt ; and again Paulus Diaconus, p. 359, 1 
 Th. : — s pro r littera saepe antiqui posuerunt, ut majo^ibus, 
 melio^ibus, la^ibus, fesiis. [Livy expressly remarks that 
 the name Furius was by certain old writers written Fusius 
 (iii. 4, 1).] 
 
 We are enabled to fix the approximate date when 
 rhotacism set in by means of two special passages. 
 Cicero, ad Familiares^ ix. 21, 2, states that L. Papirius 
 Cursor, dictator B.C. 339, was the first person to bear the 
 name of Papirius instead of the former Papirius : and the 
 Digests (1, 2, 2, 36) mention that Appius Claudius Caecus 
 (Censor in 312 B.C., and Consul in 307 B.C. and 296 B.C.) 
 changed the spelling Vale^ii and Fu.<?ii to Valerii and 
 Fu?'ii. Now, as we know that of all the words in any 
 language proper names are the least accessible to change, 
 we cannot go far wrong in assuming that rhotacism in 
 common nouns had finally set in somewhere about the year 
 350 B.C. 
 
 The apparent exceptions to the universal application of 
 rhotacism admit of several explanations. Such compounds 
 as de^uper, nisi, resedo are evidently restored forms after 
 
 - ^ 
 
 y' 
 
 O 
 
 '?) 
 
 I 
 
 9 
 
 "J 
 
 J 
 
 
 HISTORY OF THE LATIN CONSONANTS 49 
 
 the model of super, si, sedo. In other cases where s 
 between two vowels appears unchanged, we have to deal 
 with an old -ss-, which, after a long vowel or diphthong, 
 was simplified into -s- (§ 58). Thus causa, casus, divisio 
 were written and pronounced till Cicero's time, and even 
 later, caussa, cassus, divissio ; cf. Quintilian, Instit. Orat, i., 
 7, 20 : quid ? quod Ciceronis temporibus paulumque infra 
 fere quoties s littera media vocalium longarum vel subjecta 
 longis esset, geminabatur ? ut caussce, cassus, divissiones : 
 quomodo et ipsum et Vergilium quoque scripsisse manus 
 eorum docent. Finally we meet s between two vowels in 
 a number of loanwords which found their way into Latin 
 after 350 B.C., and so escaped the influence of rhotacism ; 
 c/., e.g., cisium, *a light, two-wheeled carriage,' gaesum, 
 *a heavy iron javelin,' both Celtic words; basis, 'founda- 
 tion,' nausea, ' sea sickness,' pausa, ' stoppage,' which are 
 of Greek origin; and asinus which seems to have come 
 from Asia Minor through some Thracian dialect. ^ 
 
 § 42. If we disregard words of the type of mil^s, 6s a 
 bone, etc. (for which see § 56, 4), the law holds good that 
 final s in Old Latin was dropped when standing after a 
 short vowel, and when the following word began with a 
 consonant; cf. Cicero, Orator, 48, 161: quin etiam quod 
 jam subrusticum videtur, olim autem politius, eorum 
 verborum quorum eaedem erant postremae duae litterae 
 quae sunt in optimus, postremam literam detrahebant, nisi 
 vocalis insequebatur. And in fact the dental spirant under 
 these conditions is seldom written in the oldest inscriptions. 
 The Old Latin poets also generally treated it in this 
 position as not affecting the scansion ; thus, for instance, 
 
 1 [Hehn, Kulturpjlanzeyi und Ilausthiere, p. 475, assumes the derivation 
 of asinus to be from the Hebrew athoti.] 
 
 D 
 
 
50 
 
 LATIN PHONETICS 
 
 Eunius writes ' Ancu' reliquit,' and Lucilius, * Aeserninu' 
 fuit.' When, however, the following word began with a 
 vowel, a process of liaison, as in French, set in ; i.e,^ final 
 s was separated from the previous word and connected 
 with the initial vowel of the following word; this pre- 
 vented its being dropped (cf. optimos^omnium like French 
 mauvais^esprit). Analogy, however, gradually introduced 
 regularity into the treatment of final s by restoring its 
 value in all cases alike. In Lucretius it is proved by 
 statistics that the cases in which the presence of a final 
 s before a word beginning with a consonant makes a 
 preceding short vowel long by position, are more numerous 
 than those in which it is neglected. Catullus shows but 
 one instance of final s which has not caused a lengthening 
 by position before a word beginning with a consonant. 
 This example is in Carmen, 116, 8 : at fixus nostris tu 
 dab?(s) supplicium. 
 
 The disappearance of final s in Old Latin, of which 
 we have just spoken, explains the presence of duplicate 
 forms like magis and mage, amari^ and amare (thou art 
 loved). The old pronunciation was really, e.g.y m&gis 
 elegans, but magi' saepe. Since, as we have seen 
 (§ 28), every absolutely final t became e, *magi became 
 mag6. Once this change had taken place there was 
 naturally no restoration of final s possible ; hence the 
 two parallel forms magis and mage survived side by side ; 
 similarly with amaris and amare. Analogy, however, 
 interfered with the original conditions ; the use of magis 
 and amaris soon ceased to be connected with position 
 before a vowel, and in like manner the use of mage 
 and amare was no longer connected with position before 
 a consonant; the initial letter of the following word 
 ceased to play any part in the choice of one or the other 
 
 '•> 
 
 HISTORY OF THE LATIN CONSONANTS 51 
 
 of the two forms [cf. also potis, pote ; sive = * sivis ; neve 
 = *nevisj.^ 
 
 § 43. Sometimes final s was replaced by r. Thus the 
 word labor, as it is written in Classical Latin, took the 
 place of the archaic form lahos (for the shortening of 
 the in labOr see § 31, 2); and similarly vapor replaces 
 vapo5 (cf. Quintilian, Listit. Orat., i. 4, 13 : nam ut 
 Valesii Fusii in Valerios Furiosque venerunt, ita arbo5, 
 labos, vap05, etiam . . . fuerunt). 
 
 This change must not be considered as t:ie result of a 
 regular sound-development ; the final r of the nominative 
 case labor, vapor, etc., is simply the r of the oblique cases, 
 which has extended its influence beyond its original limits 
 owing to the analogy of nomina agentis such as dator, 
 tonsor, whose r was Indo-European. It is worth noting, 
 too, that although the nominative honCr is found on an 
 inscription as early as 130 B.C., hono5 is unquestionably 
 the prevailing form in the best MSS. of Cicero, Livy and 
 Horace, and that Vergil employs exclusively the form arbo^, 
 and never arbor. 
 
 h. 
 
 § 44. From the beginning of the literary period, the 
 breathed guttural spirant h (the sound in Scotch 'loch') 
 was still merely a breathed sound in the larynx, produced 
 by the friction of the breath against the borders of the 
 vocal chords. Thus the Latin grammarians never speak of 
 the symbol A as a littera, but only as a nota aspirationis. 
 Lightly pronounced when initial, and almost inaudible 
 when medial, this breathed sound fell into complete 
 
 ^ [In the greater part of Italy, and probably in Dacia, final 5 disappeared 
 once for all from the common pronunciation in the second and third 
 centuries A.D., except in monosyllables; see Grandgent, Introduction to 
 Vulgar Latin, § 29.] 
 
52 
 
 LATIN PHONETICS 
 
 disuse in both positions at an early date. We know that 
 words beginning with h were treated in prosody in precisely 
 the same way as words commencing with a vowel ; and 
 that medial h had no power to prevent the ihotacism of 
 5 in *dishibeo, 'I separate,' which became diribeo (§ 41), 
 nor the contraction of nihil and nehemo to nil and nemo 
 respectively (§ 32). In spite of this, educated society re- 
 introduced the aspiration of this symbol under the influence 
 of orthography, and in classical times neglect to aspirate 
 the h was regarded as a sign of bad education or low origin. 
 But as this was a purely artificial pronunciation, many 
 people introduced h wrongly even into words where it did 
 not belong, cf. the well-known epigram of Catullus, in 
 which the poet derides a certain Amus who instead of 
 insidias said ^insidias (§ 35). Under these circumstances 
 an uncomfortable uncertainty crept in even into the 
 orthography, as in many cases no sure criterion was at 
 hand for distinguishing the words which rightly took the 
 aspiration from those where it was incorrect. Thus it 
 became customary to write, instead of the correct form 
 Aanser, 'a goose,' the rural and plebeian form anser.^ 
 
 On the other hand the customary orthography spelt 
 umerus as humerus adding an h which had no right to be 
 there. This explains, too, the inconsistencies in the old 
 texts, and consequently also in the modem editions, which 
 hesitate between irpex and Airpex, ' harrow ' ; erus and 
 Aerus, ' master ' ; arundo and ^arundo, ' reed ' ; olus and 
 Aolus, 'vegetable,' etc. 
 
 It is therefore not strange that the Roman grammarians 
 from the earliest times made it one of their principal 
 occupations to lay down clearly the correct facts as to 
 aspiration. These efforts, for the most part futile, may 
 
 1 [Or perhaps through analogy of anas which never had an A.] 
 
 ^ 
 
 M 
 
 •I 
 
 ♦1 
 
 I 
 
 k 
 
 \ 
 
 f 
 
 e 
 
 HISTORY OF THE LATIN CONSONANTS 53 
 
 be compared with those of French teachers who give 
 themselves much unnecessary trouble in impressing on 
 their pupils the difference between h aspirate and h mute, 
 although h aspirate has been unknown in France for at 
 least two centuries, except in some local dialects such as 
 those of Normandy and Lorraine. 
 
 Trills and Divided Consonants. 
 
 g 45. When two r's or two /'s are in close proximity 
 their tendency is to suffer differentiation ; i.e., the articula- 
 tion of one of the two sounds is modified, or actually 
 suppressed under the influence of the other. This character- 
 istic, known under the name of Dissimilation, follows certain 
 principles which are analogous to the other phonetic laws, 
 but much more difiicult to formulate. It seems, therefore, 
 best to confine ourselves here to an enumeration of the 
 actual changes attested by historical evidence. 
 
 (1) l-l>l-r: 
 
 The suffix -alls was replaced by-aris, whenever the word, 
 to which it was attached, already contained an /, cf. 
 auxiliaris, consularis, lunaris, militaris as against 
 mortaZis, navcaZis, regaZis, venaZis, and similarly pulvinar 
 (a cushion on which the effigies of the gods reclined during 
 a lectisternium) as against cervicaZ (pillow), cubitaZ (elbow- 
 cushion). These substantives are neuters from adjectives 
 in -alis, used as substantives ; for the loss of final ^ see 
 § 31, 1 ; for the shortening of a see § 31, 2. 
 
 For the same reason the suffix -crum has taken the place 
 of -clum in the w^ords fulcrum, lucrum, sepulcrum as 
 against pericZum, pocZum, vincZum. {Cf., too, lavacrum, 
 simulacrum. ] 
 
 (2) l-l>r-l: 
 
 Caeruleus, 'blue,' from *caeZuleus from caelum ; Parilia, 
 
 -ium (the festival of Pales) from PaZilia. 
 
 (3) r-r>l-r: 
 
 Zemuria, -iorum (Roman festival held on 9th May to 
 
» *' '■srr^^'i * ■■ 
 
 
 
 54 LATIN PHONETICS 
 
 propitiate the dead and the evil spirits) from *Remuria 
 according to Ovid, Fasti, v. 479 sqq. 
 
 The Vulgar Latin form pe/egrinusforperegrinus (c/. the 
 English pi/grim, and French p^^erin). 
 
 F/agrare, to cause a smell of burning, from fragrare, cf. 
 French f/airer [which means at present *to scent,' but 
 which formerly signified ' to spread a smell ']. 
 
 (4) r-r>»*-nil: 
 
 crebesco, crebui from crebresco, crebrui. 
 
 praestigiae, 'juggler's tricks,' connected with praestringo, 
 
 'I dazzle.' 
 
 Vulgar Latin propius and propietas from proprius and 
 proprietas {cf. Fr. propietaire, a common mispronunciation 
 for proprietaire). 
 
 r-r>nil-r: 
 
 Fabaris, an affluent of the Tiber in the Sabine territory 
 (Vergil, ^e7i., vii. 715), called by the Oscans Farfarus, Ovid, 
 Met., xiv. 330 ; for the treatment of Latin b as against 
 Oscan/see § 40. 
 
 Nasals. 
 
 § 46. The treatment of final m is the only point which 
 
 calls for remark. 
 
 We have four different authorities on which to base our 
 conclusions as to the value of final m in Latin : 
 
 (1) The statements made by the Latin grammarians ; 
 (2) the orthography of the inscriptions; (3) metre; 
 (4) comparison with the Romance languages. 
 
 (1) Of the passages in grammarians in which the 
 definition of the nature of final m is attempted, the 
 following three seem of the greatest importance for our 
 purpose : — 
 
 (a) Quintilian, Instit. Ch-at., ix. 4, 40: atqui eadem ilia, 
 littera (i.e., m) quotiens ultima est et vocalem verbi 
 sequentis ita contingit, ut in earn transire possit, etiam 
 
 •% \\ *' 
 
 ^i 
 
 * » ' 
 
 1 
 
 HISTORY OF THE LATIN CONSONANTS 55 
 
 si scribitur, tamen parum exprimitur, ut multum ille et 
 quantuvi erat, adeo ut paene cuiusdam novae litterae 
 sonum reddat ; neque enim eximitur, sed obscuratur et 
 tantum in hoc aliqua inter duas vocales velut nota est, 
 ne ipsae coeant. 
 
 (&) Velius Longus, vii. p. 78, 19 k [a grammarian who 
 lived in Trajan's time]: cum dico etiaiii nunc, quamvis 
 per m scribam, nescio quomodo tamen exprimere non 
 
 possum. 
 
 (c) Priscian [about the beginning of the fifth century 
 A.D.], ii. p. 29, 15 k: m obscurum in extremitate dictionum 
 sonat, ut templum; apertum in principio, ut magnus ; 
 mediocre in mediis, ut umbra. 
 
 (2) In the Old Latin inscriptions final m is very often 
 dropped. Thus we read in the sepulchral inscription in 
 saturnine verses of L. Cornelius L. F. Scipio, consul 259 
 B.C., oino for omom ( = unum); dvonoro for dvonoro??i 
 ( = bonorum) ; viro for virow ( = virum) ; aide for aidem 
 ( = aedem); but also Luciom = Lucium. This habit of 
 dropping final m was continued during the following 
 centuries on plebeian inscriptions, and indeed became more 
 and more frequent, whilst in the orthography of the official 
 inscriptions the insertion of final m was practically never 
 neglected from the second half of the second century B.C. 
 In rare cases we find a final -m replaced by -w (e.g., salvow 
 for salvowi, ta?^ for tarn). 
 
 (3) In poetry if a word ends in -im, -em, -am, -om, -urn, 
 and the following word begins with a vowel, there is 
 always elision just as if no m were there ; on the other 
 hand, these syllables are scanned as long if the following 
 word begins with a consonant ; cf. Vergil, Aen., ii. 101 : 
 Sed quid ego haec autem nequiqu«»" ingrata revolvo ? 
 
 4. The Romance languages show no trace of final -m, 
 except in a small number of stressed monosyllables like 
 
 ■'■a 
 
 
56 
 
 LATIN PHONETICS 
 
 the French riew from Tern. [The Romance languages 
 point to a loss of -m in all words but monosyllables; 
 see Grandgent, Introductioji to Vulgar Lathi, § 309.] 
 
 These different statements can be reconciled only by 
 assuming that as early as the pre-literary period of 
 Latin, final m was reduced to a mere nasalisation of the 
 vowel preceding it. As the Latin alphabet, like the 
 English, possessed no special symbol for the representation 
 of nasal vowels, spellings like viro, equo;/?, and salvo/i 
 represented three different phases in the attempts at 
 denoting nasalisation in writing. This enables us to 
 understand why Velius Longus called the sound in 
 question 'undefined,' and also why Quintilian spoke of 
 it as 'paene cuiusdam novae litterae sonus.' As nasal 
 vowels preceding a consonant were always long, it is 
 natural that scansions like 
 
 VThem quam dicunt Romaw, Meliboee, putavi 
 
 represent the normal treatment in scansion of the groups 
 -erriy -am, etc., before a word beginning with a consonant. 
 And, finally, this explains the elision of these same groups 
 before a following vowel ; for, since the nasalisation was 
 produced during the utterance of the vowei, and not after- 
 wards, the poets could not do otherwise than put these 
 groups on the same footing, as regards liability to elision, 
 with the pure mouth-vowels. 
 
 Semi-vowelfl. 
 
 § 47. In the sound combinations forming two syllables 
 i-\- any other vowel than /, and w + any other vowel than 
 «,^ after i and u respectively the semi-vowels j and v 
 developed as transition sounds. These parasitical sounds 
 
 1 [i+i was contracted to i and w+w to w (§ 32).] 
 
 p.' 
 
 ^ 
 
 (': 
 
 • 
 
 
 HISTORY OF THE LATIN CONSONANTS 57 
 
 were not commonly denoted in writing at all. Thus the 
 written form was pius, via, duo, pluit, but the pronuncia- 
 tion was pi-i-us, vi-y-a, du-iJ-o, plu-v-it. The suppression 
 of V after u in the written language was actually extended 
 to V when organically correct, as in words like flui'ere, 'to 
 flow ' (which comes from ^flug'ere (§ 39) ; conflovont ap- 
 pears on an old inscription); fruror, ' I enjoy' (from ^frug^or 
 (§ 39) ; these were written fluere, fruor. But since the 
 Latin alphabet did not distinguish between i and J, and 
 between u and v (§ 6, note iv, 2) the ambiguity of 
 spellings like IVENTA, 'youth,' which might be read 
 either as juventa or as *iventa, or of PLVIA, 'rain,' 
 which could stand for either pluvia or "^pluja, caused 
 toward the end of the Republic the restoration or intro- 
 duction of the symbol ^^ in the form V, of course, in 
 these words and in some others of the same kind as for 
 example flunus. 
 
 Note. — The appearance of the parasitic semi-vowels j 
 and V after i and u followed by vowels is fovmd also 
 in modern French, cf. plier, pronounced pli-y-er and 
 bourard. 
 
 §48. j between two vowels was pronounced i+j. For 
 this reason Cicero, in order to reconcile the spelling with 
 the pronunciation, wrote azVo, Auax, Ma//a, instead of a^o, 
 Amx, Ma/a, the way in which these words were ordinarily 
 written, cf. Quintilian, Instit. Orat., i. 4, 11 : sciat etiam 
 Ciceroni placuisse a?7o Ma?/amque geminata / scribere, and 
 Velius Longus, vii. p. 54, 16 k, in plerisque Cicero videtur 
 auditu emensus scriptionem, qui et AzVacem et Ma//am per 
 duo i scribenda existimavit. Forms like az/unt, e/ms, 
 ma/eorem, are found also in inscriptions and in a number 
 qf our oldest MSS, 
 
 ■-i 
 
58 
 
 LATIN PHONETICS 
 
 § 49. V occurring between two vowels whose acoustic 
 quality was essentially identical was regularly lost, and 
 the two vowels were contracted {§ 32). 
 
 Examples. 
 
 ditis, genitive of dii?es, from diritis; in the same way 
 we have ditior, comparative, and ditissiraus, superlative, 
 from diritior and diuitissimus. 
 
 sis, * pray,' from si ris. 
 
 latrina, from laratrina. 
 
 introrsum, retrorsum, dextrorsum, from *introt'orsum, 
 *retrororsura, *dextroi"orsum (vorsum is an archaic form 
 from versum). [Cf. our 'Hawarden' pronounced with 
 loss of w, Lindsay, S.ff.L.G., p. 14.] 
 
 The duplicate forms diritis, dititior, diiitissimus, and 
 laratrina are due to analogy. Thus the genitive dintis is 
 clearly a new formation after the analogy of the nominative 
 dires, where the v had to be maintained. On the other 
 hand, the oblique cases ditis, diti, ditem, dite have called 
 into existence a nominative dis along with dires. 
 
 § 50. In the same way v was lost in the combinations 
 -6vQ' and -ev6- before a consonant, but with this restric- 
 tion, that the second of the two vowels did not stand 
 in the final syllable, o + e were then contracted into o, 
 while e + o did not undergo contraction (cf. § 32). 
 
 contio from coientio (corentio is attested in S. C. de 
 Bacch.). 
 
 nonus from *noi;enos; novem for *noven owes its -m 
 to decem. 
 
 deorsum, seorsum from *derorsom, *sei'orsum (for the 
 shortening of the e after the loss of r, see § 27). 
 
 but 
 
 fovea (ditch) and norem, retain the v, since in the former 
 the combination -ove- stands not before a consonant but 
 before a vowel, and in the latter the second of the two 
 vowels belongs to the final syllable. 
 
 iH 
 
 ^ 
 
 '.^ <, 
 
 HISTORY OF THE LATIN CONSONANTS 59 
 
 § 5L In Vulgar Latin the disappearance of v between 
 two vowels seems to have set in on a much larger scale 
 than in the literary language. The ancient grammarians 
 state that the people said failla for farilla, paor for 
 patJor, probai for probavi. The inscriptions preserve 
 forms like paimentum for p^vimentum, dedicait for dedi- 
 carit. The Romance languages seem also to testify to 
 the loss of V in this position ; cf. Fr. paon from * paonem ; 
 peur from * paorem ; and the so-called past definite tenses 
 like chantai, prouvai, which presuppose a Latin *cantai, 
 * probai. 
 
 § 52. The semi- vowel v became a spirant from the end 
 of the second half of the first century a.d. Thus after 
 this date we find the inscriptions hesitating between it 
 and b, which also became a spirant (§ 36) [further instances 
 from inscriptions are 6iginti, b'lxit, ^otum, rene ; Valeria, 
 Mctor, bos ; bivere for rivere is very common. See 
 Grandgent, § 316]. 
 
 § 53. Some four centuries later j in its turn became a 
 spirant, with the sound dz which it had in Old French 
 and still has in modern Italian. As the Latin alphabet 
 had no special sign to represent this new soimd, we find 
 no less than four different symbols on the inscriptions, 
 i.e., z, 5, gi, di; cf. forms like Zulia = /ulia; Zanuarius and 
 G^ianuarius = t/Januarius ; /S'ustus = Justus ; Giove = Jove ; 
 corfmx = co(n);ux, found in Low Latin inscriptions belong- 
 ing to the beginning of the Middle Ages. 
 
 Double Consonants. 
 
 § 54. When in articulating a consonant special muscular 
 energy is called into play, and the interval which normally 
 passes between the time when the organs assume the 
 necessary position and when they abandon it, is prolonged, 
 
 
 <»* 
 
 ja 
 

 t; 
 
 60 
 
 LATIN PHONETICS 
 
 the ear in such a case fancies that it catches separately 
 the noise made by the closing and opening movements. 
 In this case we speak of a double consonant. In writing, 
 double consonants were represented in Latin from the 
 second century B.C. by a repetition of the symbol employed 
 for a single consonant. It is, however, specially to be 
 remarked that from a physiological point of view they 
 do not denote two separate sounds, but only the doubled 
 time of articulation. The double consonants consist 
 merely of one single articulation, strengthened and pro- 
 longed. Thus when the Roman grammarians lay it down 
 that in words like a56fiduos, siccus, currit, fa//it, two 
 successive movements of the organs of speech were 
 necessary to the production of the consonants s, c, r, /, 
 they were misled by their imperfect auditory impressions 
 and by the customary spelling. This explanation will enable 
 us without any fear of misunderstanding to continue to 
 employ the inexact but convenient- and generally used 
 term double consonant. 
 
 A. Doubling of Simple Consonants. 
 
 § 55. In a certain number of Latin words a breathed 
 stop between two vowels, when it terminated the initial 
 syllable and followed a long vowel, was doubled with a 
 shortening of this long vowel. "We may regard it as 
 certain that this doubling was due in the first place to 
 the initial stress accent of pre-literary Latin (§ 7), but at 
 the same time the process is obscure, firstly because it 
 seems to be absolutely sporadic, and secondly because 
 nearly always the original form with the single consonant 
 after a long vowel remained in use side by side with 
 the later form which had the double consonant after a 
 short vowel. This extremely delicate and complex question 
 
 sil 
 
 #•> 
 
 HISTORY OF THE LATIN CONSONANTS 61 
 
 will no doubt have to wait some considerable time for 
 
 its solution. 
 
 Examples. 
 
 ciippa., 'tub,' 'cask,' beside cuj^a (the two variants are 
 represented in French where coupe came from cuppa and 
 cuve from cupa. 
 
 Ju;?;?iter beside Juj^iter. 
 
 libera ,, libera. 
 
 littus .. li^us. 
 
 i2 
 
 hsiCC3i 
 
 muccus 
 
 >> 
 
 >> 
 
 baca. 
 mucus. 
 
 B. Simplification of Double Consonants. 
 
 § 56. Every double consonant was reduced to a single 
 ^ consonant : — 
 
 (1) After a short vowel of the initial syllable in poly- 
 syllabic words in which the second syllable was long 
 either naturally or by position. 
 
 - Examples. 
 
 canalis, derived from ca?i7ia, * reed.' 
 "li curulis, derived from curms. 
 
 disertus, perf. part. pass, of dfssero. 
 q/ella, 'a small cake,' dim. of o/a^ ,- 
 sacellus, dim. of saccus. 
 
 As we might naturally expect, analogy re-introduced 
 double consonants into a large number of words, c/, e.g. : 
 
 U gUlma, for *gaZina, on the model of gidlxis. 
 
 i?moxius for *i7ioxius, on the model of mnocuos. 
 serratus for *seratus, on the model of serra ' a saw.' 
 
 (2) Before a consonant. 
 
 Examples. 
 ^' pergo from *perrgo, *perr(e)go (per + rego : for the 
 
 syncope, see § 15), cf. the perfect perrexi. 
 
 
^^'«-m^^^y:'^^''^''^jr^.^-^^.:y:r^^^^'i^?^^,--i:.T 
 
 o 
 
 62 LATIN PHONETICS 
 
 a5piro from *as5piro, which itself stands for atspiro, 
 adspiro (ad + spiro (§ 68)). 
 
 discindo from *dis5cindo (dis + scindo). 
 
 dispicio from *disspicio (dis + specio ; for the vowel- 
 weakening, see § 10, 1, b). 
 
 disto from *di«to (dis + sto). 
 
 buL 
 
 perrumpo, dissimilis, dissocio. 
 
 New formations due to analogy: accresco, oppnmo, a/^raho, 
 on the model of accedo, oppngno, attiugo. The spellings 
 disscindo, di^^picio, etc., insisted on by certain Latin gram- 
 marians are purely artificial, and depend not so much on 
 current pronunciation as on considerations of etymology ; 
 cf. on this point Cassiodorus,^ vii. p. 205, 18 k sqq. : 
 disspicio verbum ... per duo 8 scribendum est, non per 
 unum, quoniam ex praepositione et verbo constat esse 
 compositum, quemadmodum et conspicio^ aspicio, despicioy 
 ac per hoc per duo s disspicio scribi debet, et ita dividi, 
 dis et spicio. 
 
 (3) After a consonant. 
 
 Examples. 
 
 corculum, 'little heart' (used especially as a term of 
 endearment), from *corcculom ; original form *cordculom 
 (c/. gen. of cor, cordis) > *cortculoni (§ 62) > *corcculom 
 (§ 76). 
 
 sarmentum, 'twigs,' from *sar??i7nentom; the original 
 form was *sarpmentom, belonging to the same root as the 
 verbsarpere, * to lop off,' ' prune '> *sarbmentom (§62)> 
 *sarmnientom (§ 70). 
 
 ar^i, perfect of ardeo, from *ar55i, which itself came from 
 *artsi, *ardsi (§ 68). 
 
 1 [Died at the end of the sixth century a.d. Author of De Orthographia, 
 borrowed from the works of twelve grammarians, beginning with Donatus 
 and ending with Priscian. Cf. Sandy's Hist. Class. Scholarship, vol. i. 
 p. 244.] 
 
 \ 
 
 1 
 
 -),•! 
 
 
 % ' 
 
 
 HISTORY OF THE LATIN CONSONANTS 63 
 
 senji, perfect of sentio, from *sen5si, which itself came 
 from *sentsi (§ 68). 
 but 
 siccus, su?/iwus, pe55imus. 
 
 The double consonant was restored to gain etymological 
 clearness in cases where the prefix ex- was connected with 
 a word beginning with s. Thus the forms exsatio, ex^olvo, 
 ex5omnis became, according to rule, *exatio, *exolvo, 
 *exomnis; but since the simplification of the double 
 consonant seemed likely to obscure the etymological con- 
 struction of these words, ex6-atio, ex^olvo, ex.^omnis were 
 restored, in spelling at least. For the same reason the 
 words exculpo, expolio, extruo, which represent the cor- 
 rect forms (cf. 2 and 3 above), are commonly replaced 
 by ex5Culpo, ex^polio, ex^truo. \! \ -t - c«4. v - - ->>6c 
 
 (4y When final. / 
 
 /^ 
 
 J 
 
 Examples. 
 
 es, 'thou art,' from *ess (*es-s, 2nd pers. sing. pres. indie, 
 of es-se, like ama-s, 2nd pers. sing. pres. indie, of ama-re). 
 
 fe^, *gall,' from *fe^^, cf. gen. fellis. 
 
 hoc, nom. ace. sing, neuter, from *hocc (original form 
 ♦hodce>*hocc) (§ 31, 1 ; § 76). 
 
 miles, from *miless, which itself came from *milets 
 (§ 68). 
 
 OS, * bone,' from *os5, cf. gen. 055is. 
 
 It is, however, to be remarked that even though the 
 spelling no longer exhibits any traces of final 'double 
 consonants,' we find a certain number indicated by the 
 usages of prosody. Thus Plautus scans es as long, and 
 miles as a spondee (miles), and Vergil quite commonly 
 begins hexameters with words like hoc erat (Ae?i.j ii. 664) 
 hoc illud (ib. iv. 675), hoc opus (ib. vi. 129): these scan- 
 sions testify to the pronunciation ess, miless, hocc. On 
 
 - ». 
 
 / 
 

 ■■*- » ». ,-» ' 
 
 64 
 
 LATIN PHONETICS 
 
 the other hand, it must be observed that miles forms a 
 trochee as early as the time of Ennius, and Terence never 
 scans the last syllable of words like ades and potes as long 
 except when the following word begins with a consonant. 
 
 These apparently contradictory pieces of evidence may 
 perhaps be explained in the following way. Originally, final 
 double consonants were, as a rule, simplified only before a 
 pause (i.e.y at the end of a sentence), and before a word which 
 began with a consonant ; while before words commencing 
 with a vowel they remained, as a rule, unchanged. Scan- 
 sions like esSj miless, hocc, are the last traces of this 
 original condition. Analogy effaced it by gradually bring- 
 ing about, in all cases, the simplification of the final double 
 consonants. Hence came the scansion mileSj ades^ potes 
 with the last syllable short. 
 
 It is important to note that the final s of words like es, 
 miles y OS is in Old Latin never silent like the same letter in 
 amicus y civ is j maniis^ spes, etc. (§ 42) ; i.e.y when followed 
 by a consonant it always made a preceding short vowel 
 long by position. 
 
 § 57. The double stops were simplified after a long 
 vowel. 
 
 Examples. 
 
 secubo, from *seccubo, which itself goes back to *setcubo, 
 *sedcubo (§ 76). 
 
 se^^aro, from *se^2?aro, older *setparo, *sedparo (§§ 14 and 
 76). 
 
 § 58. -ss- remained intact in every case till the end of 
 the first century B.C., but after that time was reduced to -s- 
 when following a long vowel or diphthong ; see Quintilian, 
 Inst it. Orat.y i. 7, 20 (the passage already quoted in § 41). 
 Thus the pronunciation of Cicero, and of Vergil klso, was as 
 
 « 
 
 V 
 
 f 
 
 i 
 
 HISTORY OF THE LATIN CONSONANTS 65 
 
 given below ; we may infer this not merely from the testi- 
 mony of Quintilian, but also from contemporary inscrip- 
 tions and the best manuscripts of these authors. 
 
 classis ; essQ ; gessi, perf. of gero (from *geso, § 41) ; 
 missus, perf part, pass, of mitto (§ 83) ; cassus -us, from 
 ♦cadtus (§§ 26 and 83) j divissio, from *dividtio (§§ 26 and 83), 
 caussa. 
 
 On the other hand, in Quintilian's time the pronuncia- 
 tion was ; 
 
 classis, Qsst, gessi, missus ; but casus, divisio, causa. 
 
 As regards this last example, caus^a seems to have 
 remained in use, together with cau^a, during the whole 
 of the first century a.d. This seeming anomaly is to be 
 explained by the fact that we are dealing with a term 
 belonging especially to legal phraseology which affected 
 archaisms. 
 
 In view of the linguistic facts thus stated it were to be 
 wished that modern editors would make up their minds to 
 avoid such spellings as casus, divisio, causa in texts of 
 Cicero, Caesar, Vergil, etc. Cassus, divissio, caussa are the 
 only forms that correspond to the pronunciation of the 
 republican era. 
 
 § 59. 41- was simplified to -/- ; 
 
 (1) After a long vowel, when the following syllable 
 
 contained an i. 
 
 (2) After a diphthong. 
 
 Examples. 
 
 (1) mi/ia, nom. and ace. plur. of vollle. 
 stiZicidium, 'dripping,' )( sti/^a, 'a drop,' from * stir-la 
 (§ 73), ef. stiria, ' a drop,' or ' an icicle.' 
 vicious, ) ( viZZa. 
 steZio, * spotted lizard,' ) (steZZa, 
 
 £ 
 
66 
 
 LATIN PHONETICS 
 
 The length of the root-vowel in mille, villa, stella is 
 vouched for by the evidence of the Romance languages. 
 
 (2) auZa, ' a pot,' from au7/a (attested by the oldest 
 Plautine MSS.) )( the Vulgar Latui duplicate form olla, 
 (§ 24). 
 
 cae/um, ' chisel,' from * cae//um, from * caedlom (§ 71) 
 belonging to the root of caedo, ) ( ra/Zum, ' instrument 
 for scraping ploughshare,' from *radlom (§ 71), 
 
 § 60. -mm- after a long vowel or diphthong was reduced 
 
 to -7W-. 
 
 Examples. 
 
 gluma, *husk of corn,' from *glu?n??ia, *glubma ; glubo, 
 ' I peel ' (§ 70). 
 
 ra??ientum, ' shavings,' * chips,' from *ra??imentom, 
 *radraentom (§ 70). 
 
 cacTTientum, ' cement,' from *caemwentom,*caed;;ientom; 
 caedo (§ 70). 
 
 Kl 
 
 COMBINATIONS OF CONSONANTS 
 
 67 
 
 COMBINATIONS OF CONSONANTS. 
 
 A. Groups of Two Consonants. 
 
 1.— Assimilation. 
 
 § 61. In any language when two consonants of different 
 acoustic quality come into contact, there is a tendency to 
 suppress, or at least modify, the transition from the first to 
 the second by an entire or partial levelling of their 
 respective natures. This linguistic phenomenon is known 
 under the name of assimilation. Assimilation can affect 
 breath and voice {i.e., if two consonants, one breathed, the 
 other voiced, are in combination, they are likely to become 
 both either breathed or voiced). It affects also the manner 
 and the place of articulation. It is progressive or 
 regressive, according as to whether it is the first or second 
 consonant from which the process of levelling proceeds, or, 
 in other words, whether the assimilating consonant is the 
 first or the second. In Latin, regressive assimilation was 
 much more common than progressive. 
 
 (a). Breath and Voice Assimilation. 
 Kegressive Assimilation. 
 
 § 62. Every stop or spirant becomes breathed before a 
 
 breathed stop or spirant, and voiced before a voiced stop 
 
 or spirant. 
 
 Examples. 
 
 (1) actus (for the lengthening of the a, see § 26), perf. 
 part. pass, of agro. 
 scri^tus, perf. part. pass, of scri6o. 
 
68 LATIN PHONETICS 
 
 rexi {i.e. recsi), perf. of rego. 
 
 nupsi, perf. of nubo. 
 
 (2) aftduco, ) ( a^erio. 
 
 o6do, ) ( operio. 
 
 se^mentum, belonging to the root of seco. 
 
 § 63. In other cases the results of this law are latent, 
 i.e.y obscured by further changes. Compare : — 
 
 *clau<si, perf. of clauc^o, which became claussi (§ 68), 
 and from the Augustan period, clausi (§ 58). 
 
 *su6mos superlative, 'the highest,' from *supmos {c/. the 
 comparative suj;erior), which became summus (§ 70). 
 
 *isdem, 'the same,' from *isdem, which became idem 
 (§ 25, 1 and § 86). 
 
 For further examples see §§ 66 sqq. and § 86. 
 
 § 64. If in ordinary spelling a voiced consonant precedes 
 a breathed, as in words like o^tineo, su^tilis, ple^s, ur6s, the 
 violation of the rule is only apparent. A considerable 
 mass of evidence from the Roman grammarians establishes 
 the fact that these words were always pronounced qptineo, 
 sujotilis, ple/)s, uvps; cf, for example, Quintilian, Instit. 
 Oi'at.^ i. 7, 7 : quaeri solet in scribendo praepositiones 
 sonum quem junctae efficiunt, an quern separatae observare 
 conveniat, ut cum dico ohtinuit (secundam enim b litteram 
 ratio poscit, aures magis audiunt p). By ratio, of course, 
 we must understand analogy : o6tineo, su6tilis were 
 written after the analogy of o6duro, su6dolus, and ple6s, 
 ur^s after the analogy of their oblique cases ple^is, ple^i, 
 ur^is, ur^i, etc. 
 
 Progressive Assimilation. 
 
 § 65. Its operations are always latent. 
 
 Under the influence of a preceding trilled or divided 
 consonant, a voiced spirant was substituted for a breathed 
 spirant. 
 
 COMBINATIONS OF CONSONANTS 
 
 69 
 
 { 
 
 r 
 
 Examples. 
 
 *fer;;e (*to carry'), from *ferse (compare esse, 'to be'), 
 which became ferre (§ 74). 
 
 *veke ('to wish'), from *velse, which became velle 
 (§ 74). 
 
 *fer2e and *veke are not attested forms, but the physio- 
 logical analj'sis of pronunciation proves that the combina- 
 tions -rs- and -Is- could not become -rr- and -U- without a 
 previous change of the breathed s into the voiced z, 
 
 (b) Assimilation of the Manner of Articulation. 
 
 § 66. Assimilation of the manner of articulation was 
 very common in Latin. Hence we can give only a 
 selection of the most characteristic examples, leaving the 
 student to multiply the number in the course of his read- 
 ing. It should be noted also that it is often impossible 
 to determine the constituent elements of a combination 
 assimilated from this point of view without the help of 
 the other Indo-European languages. Thus -//- arises from 
 -dl-y -nl-y -rl-j -Id; -In-, -Is-, and so it is only the com- 
 parison of the German ffals that enables us to recognise 
 that the Latin collum (' a neck ') is developed from ^colsom.'^ 
 
 If one of the two consonants in combination was voiced 
 and the other breathed, or vice versa, the assimilation of 
 the manner of articulation was preceded by that of their 
 acoustic quality, according to the laws formulated in § 62. 
 The double consonant that resulted from the assimilation 
 of the manner of articulation was reduced to a single 
 consonant in all the cases provided for by the formulae in 
 § 56 sqq. It should be added, however, that apart from 
 a few rare exceptions the double consonant, which was 
 simplified after a short vowel in polysyllabic words with 
 a long second syllable (§ 56, i.), was re-established by 
 analogy. 
 
 1 See Giles, § 184. 
 

 70 LATIN PHONETICS 
 
 Regressive Assimilation. g( 
 
 Stop + Spirant. 
 
 § 67. In combinations formed of a labial, dental or 
 guttural stop followed by /, the stop is changed to /. 
 Thus here the assimilation of the manner is combined 
 with that of the place of articulation ; the three kinds ^; 
 of stops not being replaced by corresponding spirants but 
 changing all to /. 
 
 Practically there are no examples save for the group 
 
 'Pf'^ -df-. -C/-- 
 
 _^f-> ^.^f. I >/f. 
 
 Examples. 
 
 q/Bcina (a workshop) from o^(i)ficina (opificina is found 
 in Plaiitus Miles, 880) ; compare also opifex. 
 
 o/fero from * opiero. 
 
 a/fero from arffero, the intermediate stage being *atfero. 
 
 e/fero from ecfero (ecferet is found in Plautus Aulularia, ^ 
 664, ecfari in a quotation from Ennius in Cicero, De 
 LegibuSi iii. 9 ; the triple form of the preposition e-, ec-, 
 ex- corresponds exactly with a-, ab-, abs-). 
 
 Restorations due to analogy : o^fero, ac?fero, o^fundo, 
 ac?figo, which are duplicate forms of o/fero, a/fero, 
 o/Yundo, a/figo. 0/fundo and a/figo in their turn are i^ 
 re-established by analogy of o/fero, a/fero, and other 
 words of the same type. The only regular forms are 
 *ofundo, *afigo (§ 56, i.). The pronunciation of obfero, 
 adfero, obfundo, adfigo was, of course, o/>fero, a/fero, 
 qpfundo, a^go ; the spellings obfero, adfero, obfundo, 
 adfigo correspond precisely with obtineo, subtilis, etc. ^ 
 
 (§ w). 
 
 V 
 
 I 
 
 ! • 
 
 *• « 
 
 COMBINATIONS OF CONSONANTS 
 
 71 
 
 § 68. A combination of a dental stop followed by 5 
 became ss. 
 
 ds 
 
 > tsf 
 
 > ss 
 
 Examples. 
 
 concu5si, perfect of concu^io, from * concufsi. 
 messui, perfect of me^o, from * me^sui. 
 assequor, from arfsequor, the intermediate stage being 
 * assequor. 
 
 assum, from * a^sum, from ac?sum. 
 
 Double -ss- was reduced to single -s- : — 
 
 (1) After a short vowel in the first syllable of poly- 
 syllabic words where the second syllable was long naturally 
 or by position (§ 56 (1)). 
 
 There are no examples ; analogy everywhere introduced 
 the double -ss- and effaced the results of this law (see 
 below). 
 
 (2) In a final position (§ 56 (1)). 
 
 Examples. 
 
 miles from * mile^s (compare gen. milids from * miletis 
 (§ 10, i., b)), miless. 
 
 pes from *perfs (compare gen. perfis), *pe^s, *pess. (For 
 the long vowel of the nom. pes, see § 12.) 
 
 (3) After a long vowel or diphthong, from the end of 
 the first century B.C. (§ 58). 
 
 Examples. 
 
 lusi, perfect of \udo, from *lurfsi, *luts\, Imsi. 
 plausi, perfect of plaudo, from *plaurfsi, *plaufsi, 
 plauasi. 
 
 Restorations due to analogy : a^fsequor, ac?sum, duplicate 
 forms of assequor, assum; a^^fsigno, assisto which have 
 supplanted the regular forms *asigno, *asisto ((1) above), 
 
 
 -4irt 
 
 i 
 
i 
 
 72 
 
 LATIN PHONETICS 
 
 Stop + Nasal. 
 
 COMBINATIONS OF CONSONANTS 
 
 73 
 
 § 69. Under the action of a following n, each kind of 
 stop was changed into the nasal produced by the same 
 organ. Thus : — 
 
 'bn- > -mn- 
 
 -pn- > -7nn- passing through -^n- 
 
 -dn- > -nn- 
 
 -tii' > -7in- passing through -dn- 
 
 -gn- > -hn- 
 
 -cn- > -fin- passing through -^n- 
 
 EXAMPLES. 
 
 sca??mum (* a stool,' ' a bench ') from *sca5nom (compare 
 the diminutive scaJellum). 
 
 Sawnium (the district in Central Italy whose inhabitants 
 were descended from the Sabines) from *Sa6niom (which 
 belongs to the same root as SaMni, Sa6elli, 'the Sabines'). 
 
 somnus from *sopnos {cf. so/>or, 'deep sleep'), by an 
 intermediate stage *so6nos. 
 
 annoto from arfnoto. 
 
 penna from *petna (which belongs to the root of peto, 
 ' I direct myself towards,' the oldest meaning of which 
 was ' I fly '), through an intermediate stage *pe6?na. 
 
 As for the group -g?i-j representing partly an older -cn- 
 (cf, § 62), writing has not been able to reproduce its 
 development into -hn-, because, as we have seen (§ 6, 
 note 3), the Romans did not possess a special letter to 
 denote the nasal guttural. Thus they continued to write 
 -gn- while pronouncing -hn-. 
 
 Compare : — 
 
 linnum, written lignum, ' firewood ' ; primitive form 
 *le^num (§ 16). 
 
 dinnus, written dignus ; primitive form *decnos (§§ 16, 
 62) ; intermediate stage *de5'nos. 
 
 > I 
 
 ♦ ^'' : 
 
 The proof that in lignum and dignus -gn- was pro- 
 nounced -nn- is provided by the radical t of these words, 
 which comes from an original e, the change of ^ to ^ in 
 a closed syllable being the rule before a nasal guttural 
 (§ 16). 
 
 Restorations due to analogy : a6nuo, a^nego, which 
 have almost entirely ousted the regular forms amnuo, 
 a??inego ; adwoto, a duplicate form of awnoto ; awnecto, 
 awnitor for *anecto, *anitor (§ 56 (1)). 
 
 § 70. Omitting the combination guttural stop + m, the 
 treatment of which has not yet been established, all 
 the stops followed by m became m. This development 
 presupposes the earlier change of d and t to b and p in 
 consequence of an assimilation of the place of articulation. 
 (§ 78). Hence :— 
 
 -bm- \ 
 -dm- > -bm- ) 
 
 -pm I 
 -tm- > -pm- J 
 
 > -mm- 
 
 > 
 
 -mn- passing through -bm- 
 
 For -tm-, it is true, we have no examples, but it is 
 beyond doubt that this combination developed in the 
 way indicated wherever it presented itself 
 
 Examples. 
 
 awimoveo from ac^moveo. 
 
 su7;inius from *su&mos from *su;?mos (§§ 62, 63). 
 
 Double -mm- was simplified later to -m- (§§ 56 (1), 60) 
 after a short vowel in the initial syllable in polysyllabic 
 words whose second syllable was long by nature or 
 
■^-'^ '•?^'^- 
 
 
 
 
 74 
 
 LATIN PHONETICS 
 
 COMBINATIONS OF CONSONANTS 
 
 75 
 
 position, and in all words where it followed a long vowel 
 
 or diphthong. 
 
 Examples. 
 
 omitto from *oj5mitto {op being the original form of the 
 preposition ob; § 62 (2)), *o6mitto, *owmitto. 
 
 gluraa ('husk ') from *glufema, *glu?72ma (glu&o, 'peel'). 
 
 caementum ('hewn stone') from *caerfmentom, *caem- 
 mentom (caec?o, ' break '). 
 
 Restorations due to analogy : a(7moveo, duplicate form 
 of a?72moveo (see above); su^mitto, duplicate form of 
 suwmitto, which is itself of analogical origin, the correct 
 form being *sumitto (compare omitto above). 
 
 Stop + Divided Consonant. 
 
 § 71. The voiced dental stop d followed by the divided 
 consonant / is assimilated to it. 
 
 -d\- > -II- 
 
 EXAMPLES. 
 
 aZloquor from ac^loquor. 
 graZlae ('stilts') from *gra££lae (gradior). 
 radium (instrument for scraping earth off a ploughshare) 
 from *rac?lom (rarfo). 
 se/la from *sed\a. (sec?eo). 
 
 After a diphthong double -//- is simplified (§ 59 (2)). 
 
 Example. 
 
 caelum (a sculptor's chisel) from *c&Gd\om. > *caeflura 
 (caedo). 
 
 Restorations due to analogy : ac^loquor, duplicate form 
 of aZloquor; a</latus, perf. part. pass, of adfero, afFero 
 (§ 67), duplicate form of a/latus, which in its turn analogy 
 substituted for * alatus (§ 56 (1)). 
 
 ♦ '4 
 
 ' 
 
 ^ 
 
 
 
 
 ^A 
 
 Nasal + Trilled or Divided Consonant. 
 
 § 72. Nasal w is assimilated to a following trilled or 
 divided consonant. 
 
 -nr- > -rr- 
 -n\- > -II 
 
 Examples. 
 
 (i.) corripio, from *co7irapio (with umlaut, § 10, i. c). 
 irrevocabilis, from inrevocabilis. 
 (ii.) colloquium, from conloquium. 
 
 vi^um (' a sup of wine '), from *vin(o)lum (diminutive of 
 vinum with syncope, § 15). 
 
 Restorations due to analogy : iwrevocabilis, conloquium, 
 duplicate forms of i?Tevocabilis, co/loquium (see above) ; 
 corrumpo, i//atus (perf. part. pass, of infero), which have 
 replaced the regular forms corumpo and "'^ilatus (§ 56 (1)). 
 Of corumpo there remains a last trace in Lucretius, vi., 
 1135: 
 
 An coelum nobis ultro natura c5ruptum. 
 
 Trilled + Divided Consonant. 
 
 § 73. A trilled followed by a divided consonant is changed 
 by the latter into a divided consonant. 
 
 -r\- > -II- 
 Examples. 
 
 ageHus, from *agerlos (diminutive of ager). 
 peZlicio, from *perlacio (with umlaut, § 10, i. c). 
 satuZlus ('satisfied') from *saturlos (derived from satur). 
 
 Restorations due to analogy : perlioio, duplicate form of 
 pe/licio (see above), perluceo, duplicate form of pe/luceo, 
 which is itself substituted by analogy for *peluceo (§ 56 (1)). 
 
i^^^&f?*! ■■J'^i^^^^^^f^^^'t'^riA^^^^f^^^^^^W^f^^^^'. 
 
 76 LATIN PHONETICS 
 
 Progressive Assimilation. 
 
 Trilled or Divided Consonant + Spirant. 
 
 § 74. A trilled or divided consonant assimilates a follow- 
 ing * after having first changed it into z (§ 65). 
 
 -Ts- > -rr- 
 -Is- > -1/- 
 
 EXAMPLKS. 
 
 ferre from *fer2e, from *ferse (compare esse). 
 
 velle from *vehe from *vel5e. 
 
 To the same category belong words like col/um, torreo, 
 verres which come from *colsum, *torseo, * verses. These 
 original forms, however, are proved only by a comparison 
 with the other Indo-European languages (§ 66). 
 
 To this rule there are exceptions, e.^., words which show 
 the combinations -rs- and -Is- without assimilation. Com- 
 pare for example arsi, farsi, mubi, the perfects of ardeo, 
 farcio, mulceo ; pulsus, perf. part. pass, of pello. Of course 
 this different treatment of the same combinations of 
 consonants only apparently violates the principle of the 
 constancy of phonetic laws. The formulae of these laws, 
 in fact, apply only to sounds or combinations of sounds of 
 perfectly identical nature. Now in arsi, the combination 
 -rs- comes from -rss- (§ 89) ; farsi and mulsi stand for 
 *farcsi, 'mulcsi (§ 90), and the simplification of the com- 
 binations -res and -Ics- to -rs- and -Is- came only after the 
 assimilation of original -rs- and -Is- to -rr- and -//-. Finally, 
 pulsus is an analogical creation that has replaced the 
 regular form *pultus. Of this there remains a last trace 
 in the verb pultare (' strike ') used by Plautus, Captivij 832, 
 cf. also Quintilian, Instit. Orat., i. 4, 14: nam inertare ei 
 pultare dicebant. Pulsus was substituted for *pultus at a 
 
 COMBINATIONS OF CONSONANTS 
 
 77 
 
 # ^ 
 
 ^ 
 
 time when the assimilation of original -rs- and -Is- was 
 (. equally an accomplished fact. None of these forms thus 
 corresponds exactly with *ferse or *velse. 
 
 (c.) Assimilation of the Place of Articulation. 
 
 § 75. Assimilation of the place of articulation was always 
 
 C regressive. All the remarks made above (§ 66) with regard 
 
 to the assimilation of the manner of articulation are true of 
 
 this process also. The student must therefore refer to them 
 
 before beginning to study the following laws. 
 
 Assimilation between Stops. 
 
 § 76. Before a guttural stop every dental or labial stop 
 is changed into a guttural stop; dental stops are also 
 assimilated to a following labial stop. The following table 
 summarises these changes : — 
 
 •-^i^ % 
 
 (1) 
 
 -tg- > -dg 
 
 -hg 
 .pg. > -hg 
 
 
 > -gg- 
 
 -do- > 'tc- 
 
 -pc- 
 
 -hc >-pc- 
 
 ' > -cc- 
 
 -dci' 
 
 (2) 
 
 -pq- . i 
 -dq^- > -pq ') 
 
 -dbA 
 -th' > -dh-J 
 
 ■dp- > -tp'J 
 
 > *-q* q^-, which became -cq^ - (§ 82). 
 
 > -6b- 
 
 > -pv- 
 
> 
 
 78 
 
 LATIN PHONETICS 
 
 Some of the instances of assimilation shown in this 
 table rest, it is true, only on analogical induction, as 
 there are no examples. The following are attested his- 
 torically : — 
 
 aggero, from ac?gero. 
 
 oggevo, from 'opgero, o6gero. 
 
 siccus, from *sii(i)cos (which belongs to the same root as 
 sids). 
 
 succido, from *suj5cado (with umlaut, § 10, i. c). 
 
 quicquam, from qui(;?quam, *qui^quam. 
 
 ocquinisco ('stoop'), from *o;?quinisco (compare con- 
 quinisco ; the simple verb is not used). 
 
 quijope,from *quifl?pe (quid, neuter of quis -f- pe, a particle 
 found also in nempe), *qui^pe. 
 
 A double guttural (i.e., -cc-, -pp-, etc.) was simplified 
 in a final position (^ 56 (4)), and in a medial syllable after 
 a long vowel (§ 57). 
 
 Examples. 
 
 hoc, from *hof?ce (*hod, neuter, like id, + ce, a demon- 
 strative particle ; old inscriptions have honce = hunc, 
 hance = hanc, etc. ), *hotce *hocce, and with the loss of the 
 final vowel (§ 31 (1)), hocc. (On this form, see § 56 (4)). 
 
 secubo (* I lie alone or apart '), from *sec?cubo (the prefix 
 sed denoting separation, c/. seditio), *se^cubo, *seccubo 
 (§ 57). 
 
 separo, from se^Zparo, *se^paro, *se^paro (§ 57). 
 
 Restorations due to analogy: ac?gero, obgero, qmd- 
 quam, duplicate forms of a^gero, o^gero, quicquam ; 
 ac?bibo, which has ousted *Miho doubtless owing to the 
 latter's ambiguity ; iccirco for "^icirco (§ 56 (1)), arising 
 from ic^circo, which was re-established and existed as a 
 duplicate form along with iccirco. 
 
 # Ji 
 
 f) 
 
 \ 
 
 « 
 
 COMBINATIONS OF CONSONANTS 
 
 79 
 
 Assimilation between a Stop and a Spirant produced 
 
 by Different Organs. 
 
 § 77. Before the labial spirant / every dental or guttural 
 stop was changed into a labial stop. 
 
 This assimilation is latent because the labial stop was later 
 changed into a spirant owing to assimilation of the manner 
 of articulation (§ 67). 
 
 Assimilation between a Stop and a Nasal produced by 
 
 Different Organs. 
 
 § 78. Before the labial nasal m, every dental stop was 
 changed into a labial stop. 
 
 As in § 77, this assimilation is latent, and its results 
 altered by a secondary assimilation of the manner of articu- 
 lation (§ 70). 
 
 Note —In this instance the reconstruction of the latent 
 assimilation of the dental stops to labial stops before m is 
 based partly on the physiological difficulty of a direct change 
 from -dm- and -tm- to -mn-, and partly on the very pro- 
 nounced tendency to labialise a dental stop followed hym, 
 which is seen in other languages as well as Latm. Thus 
 in the Swiss pronunciation of modern German, written 
 words like Bacimeister (' bath-attendant '), Mitoensch (' fel- 
 low-creature '), we seem to hear Bai?meister, Mii^mensch. 
 
 Assimilation of Dental Spirant s to Labial Spirant /. 
 § 79. The dental spirant s of the prefix dis- is assimilated 
 
 to a following/; 
 
 -rf- > ./f- 
 
 Compare : di/f6ro, from *disf6ro. 
 
 di/flcilis, from *di«facilis (with umlaut) 
 
 (§ 10, i. c). 
 
 ^1 
 
^^^f^e^F^-r9=v?4'Ss ■ .^^^^ 
 
 
 80 
 
 LATIN PHONETICS 
 
 Restorations due to analogy: di/fido, di/fundo, which 
 have replaced *di/ido, *di/undo, the only regular forms 
 (§ 56 (1)). 
 
 Assimilation between Nasals and Stops, 
 
 § 80. Only the corresponding nasal, i.e., formed by the 
 same organ as the stop, can precede a stop. 
 
 Examples. 
 
 The prefixes c5m- (comes), in- (ineo) and the negative 
 particle in- (in-utilis) end in m before a labial stop, in n 
 before a dental stop, and in n (written n owing to the lack 
 of a special letter to denote the nasal guttural (§ 6, note 3)) 
 before a guttural stop ; compare :— 
 
 (i.) compono, contexo, concenatio pronounced concenatio. 
 
 (ii.) imbuo, induro, ingenues pronounced ingenues. 
 
 (ill.) imbellis, i/itactus, inquietus pronounced iwquietus. 
 
 Moreover, the change from m to n before a dental stop 
 and to h before a guttural stop is seen in the following 
 cases : — 
 
 eu?idem, eandem (eum, earn + dem). 
 
 eorundem, earundera (eoru?/i, earu7?i + dem). 
 
 quandiu (quam + diu). 
 
 septendecim (septe?;i + decern). 
 
 clanculum (adv. 'secretly' and prep. * without the 
 knowledge of,' found in Plautus and Terence) pronounced 
 clanculum ; cf. cl&m, which has the same meanings. 
 
 tunc, pronounced tunc (tu7?i + enclitic ce (§ 31 (1)). 
 
 tanquara, pronounced taAquam (taw + quam). 
 
 Restorations due to analogy: eumdem, eawdem, 
 eoruwdem, earuwidem, quawdin, septemdecim, tawquam, 
 duplicate forms of euwdem, etc. ; suwptus (with develop- 
 ment of a parasitic p (§ 85)), perf. part. pass, of su?wo 
 for suwtus. The latter form, however, existed in Vulgar 
 
 " 
 
 ^ 
 
 I 
 
 COMBINATIONS OF CONSONANTS gi 
 
 Latin and survives in certain Romance forms. It 
 IS, moreover, probable that the restoration of the 
 etymological forms eu^wdem, etc., was at first purely 
 literary, that is to say, the same people who wrote with 
 the m none the less continued to pronounce the words 
 euwdem, etc. 
 
 Assimilation of the Nasal Dental to the Nasal Labial. 
 
 § 81. The nasal dental n is assimilated to the nasal 
 labial m wherever the preposition or negative particle in- 
 precedes a word beginning with m. 
 
 nm > mm. 
 
 Examples. 
 
 immolo from inmolo (properly, « I sprinkle the head of 
 the victim with the sacred flour called mola). 
 i?/imerito from irimerito. 
 
 Restorations due to analogy: inmolo, inmerito, 
 duphcate forms of i^wmolo, i^wmerito ; iwmuto, immitis, 
 the phonetic forms being *imuto, ^imitis (§ 56 (1)). 
 
 II. Treatment of g^ and gv before a Consonant. 
 
 § 82. Before a consonant the labialised guttural stops 
 g"^ and 2- (written gu and gu (§ 39)) lost their labial 
 appendix and were changed into pure gutturals g and c. 
 Moreover, if the following consonant was breathed— and 
 that is true of all the examples that can be adduced of 
 the combination g^ + consonant— the voiced g became 
 a breathed c (§ 62). Hence :— 
 
 g, q^ + consonant > c + consonant, 
 
 F 
 
 vll 
 
,p^' "? ^'' 
 
 •) 
 
 g2 LATIN PHONETICS 
 
 Examples. 
 
 extincsi, written extinxi, perfect of exi'mguo. 
 nks, written nix, )( the genitive nivis that goes back to 
 * ni^ns (§ 39). 
 
 unctio, )( ungfuentura. 
 assecla ('a hanger-on'), )( &sseq(ior. 
 coctiis, perf. part. pass, of co^uo. 
 delictum, ) ( delin^wo. 
 This law explains also ac and nee, duplicate forms of 
 at^./e and ne^.^e, and used when the following word begins 
 with a consonant. After the loss of the final e of atque 
 and neque (§ 31 (1)) the labialisation of qu disappeared 
 under the influence of the following consonant. Hence 
 ♦ate and nee; *atc was then assimilated into *acc (§ 76), 
 and finally the double cc was reduced to a single c 
 (§ 56 (4)). 
 III. Treatment of the Combination Dental Stop + t 
 § 83. From the Indo-European epoch the combination 
 of a dental stop and t gave rise to the development of 
 a medial parasitic s. Hence the combination tst, which 
 in Latin became ss by a double assimilation. 
 
 d^ t + t > tst > ss. 
 Examples. 
 cas5us from *cMtns (carfo ; for the lengthening of the 
 radical vowel in this word and the two followmg, see 
 
 §26). 
 
 divissio from *divirfdo (divin?o). 
 
 essiis, perf. part. pass, of e^o (' I eat ') from *edios, 
 
 plaus^us, perf. part. pass, of plaurfo, from ^plauc^^os. 
 
 Lsus, perf. part. pass, of sedeo, from ^sedtos (on the 
 probable reason why the radical vowel in sessus is not 
 lengthened, see § 26). 
 
 messis from *me«is (me^o). 
 
 passus sum, perfect of pador, from ♦paWos sum. 
 
 quassus, adj. old perf. part. pass, of qua^o, from *qua«os. 
 
 V 
 
 f 
 
 '} 
 
 COMBINATIONS OF CONSONANTS 83 
 
 After a long vowel and diphthong the double ss was 
 reduced to a single s from the end of the first century 
 B.C. Cicero still pronounced and wrote ca^^us, etc., but in 
 Qumtihan's time these words were spelt casus, etc. 
 (§ ^S)- 
 
 § 84. In a certain number of cases the combination 
 dental stop + t was formed afresh after the development 
 traced above which ended well before the historical period. 
 The result of this second evolution was not now -ss- but 
 -«-. This happened especially when the prefix ad- was 
 combined with a verb that began with t ; cf. :— 
 Aftenuo from a,dtenuo. 
 
 a«uli, perfect of affero (assimilated from adfero, S 67) 
 from arf^uli. ' ^ ' 
 
 ^ The prefixes attached to verbs were in fact originally 
 independent words, and this character had not yet been 
 lost completely iu Plautus' time ; cf. Trmummus, v. 833 : 
 dixtraxissent disque tulissent . . . 
 
 Thus the combination dt- in adtenuo, adtuli, was too 
 recent to be affected by the law that caused for example 
 the change of *sedtos to sessus (see above). Hence we 
 get attenuo, attuli. It should be added that by the side 
 of a^j^enuo, a«uli, restored duplicate forms, ac/^enuo, B.dtxi\i 
 are also found in inscriptions and manuscripts. 
 
 IV.-Development of a parasitic sound in certain 
 combinations of two consonants. 
 
 § 85. Between the two elements of the combinations sr 
 (as far as it represents an earlier dtr, see §§ 56 (2), 83), 
 ml, ms, mt, a parasitic transitional sound developed, which 
 in the case of 5r was a t, and in that of ml, ms, mt^p 
 
 
 ■^^ 
 
 m 
 
 mMfrmm 
 
84 
 
 LATIN PHONETICS 
 
 COMBINATIONS OF CONSONANTS 
 
 85 
 
 (i.) sr > s^. 
 (ii ) ml > mp\. 
 ms > mps. 
 int > mpt 
 
 Examples. 
 
 (1) clausftnim, from *clausrom, ♦claudtrom (claudo). 
 ras^rum, from *rasrom, *radtrom (rado). 
 tons^rix, from *tonsrix, *tondtrix (tondeo). 
 
 (2) exem;?lum, from *exemlom (properly ' that which is 
 taken out as a sample,' from *exemo, which became 
 
 eximo (§ 10, i. &)). ^ ^ 
 
 com;>si, demi^si, prompsi, sum;)si, perfects of como, 
 
 demo, promo, sumo, from *comsi, Memsi, •promsi, sumsi 
 
 (c/., for example, dixi, i.e. , dicsi, perfect of dico). 
 oom;)tus, dem;?tus, prom;>tus, sumi?tus, from comtus, 
 
 ♦demtus, *promtus, *sumtus. 
 
 In popular speech a parasitic p developed also in the 
 combination -mn-, fts is shown by the spellings, such as 
 autumjonus, contem;?no, som;?nus, which are found fre- 
 quently in vulgar texts. Educated people absolutely 
 eschewed this pronunciation. 
 
 Note.— (i.) *comtus, Memtus, *promtus, ♦sumtus ought 
 to have become *contus. Mentus, *prontus *suntus (§ 80) ; 
 but the VI was restored by analogy. 
 
 (ii.) If the parasitic consonants were relatively few in 
 Latin, French, on the other hand, is rich in them, cf. 
 O.F. ances^re, es^re from Latin antecess(o)r, *ess(e)re ; 
 mod. Fr. com&ler from Latin cum(u)lare, nom&re from 
 num(e)rum, ponrfre from pon(e)re, etc. 
 
 v.— Loss of the first element of a combination of 
 
 two consonants. 
 
 § 86. Before a voiced consonant, s changed to z (§ 63), 
 and then was lost with a compensatory lengthening of 
 the preceding vowel if the latter was short (§ 25 (1)). 
 
 > 
 
 1^ 
 
 i. 
 
 Examples. 
 
 idem, from *rsdem (is + dem), which became first *Lcdem. 
 
 judex, from *ju5dex (' quod jus dicat ') ; -dex for -dix by 
 analogy of the second element of compounds like auspex, 
 opifex, etc., according to the proportional formula, aus- 
 picis, opificis: auspex, opifex = j udicis :x), 'jusdex. 
 
 tredecim, from *tresdecim, *tre2decim. 
 
 prelum (' wine-press'), from *preslom {cf. pres-si, perfect 
 of premo), *pre2lom. 
 
 primus, from *prismos {cf. prisons), *prizmos. 
 
 egenus, from *egesnos {cf. egestas), *ege2nos. 
 
 pono, from *pos(r)no (§ 25 (1)), *po:;no. 
 
 Cf. also diduco, digero, dilanio (* I tear in pieces '), 
 dimitto, dinosco, as against distorqueo, discedo, dispone, 
 dissocio. 
 
 Restorations due to analogy : ejusdem, quibusdam, after 
 ejus and quibus. 
 
 § 87. Before / and 5 the nasal dental n lost its stop 
 character at an early date; the vibrations set up in the 
 glottis during this sound were then attracted to the pre- 
 ceding vowel, lengthening it by way of compensation when 
 it was short (§ 25 (2)). 
 
 The loss of n in this position is proved (i.) by spellings 
 like iferos, cesor, cosol, found frequently on archaic in- 
 scriptions ; (ii.) by the abbreviation cos. for consul ; 
 (iii.) by the testimony of Quintilian (§ 25 (2)) ; (iv.) by 
 * inverse spellings ' in vulgar texts, e.g.^ occawsio, thew- 
 saurus for occasio, thesaurus ; (v.) by the Romance 
 languages, q/*., for example, the French Spouse, mois, toise 
 ' fathom '), which go back to sposa, mesem, tesa. But at 
 an early period etymological considerations caused the 
 restoration of w, first in the spelling, and then, owing 
 to the instinctive popular tendency to harmonise the 
 spoken with the written language, partially at any 
 
 ■srs 
 
86 
 
 LATIN PHONETICS 
 
 rate in pronunciation. This happened especially in the 
 case of compounds with con- and in-. Hence the French 
 cowseil, e;jfant, ewsemble, that develop from consilium, 
 infantem, iwsimul. 
 
 B. Combinations of Three Consonants. 
 
 § 88. Combinations of three consonants were reduced 
 most frequently to two consonants, but occasionally even 
 to one. The simplification to two consonants was due — 
 
 (1) To the assimilation of two consonants of the com 
 
 bination, after which the double consonant was 
 made a single one. 
 
 (2) To the loss pure and simple of a consonant under 
 
 conditions formulated below. 
 
 The simplification to a single consonant was due {a) to 
 the combination of the one or other of these two causes 
 with the loss of an s, or {h) to the loss of ns before a voiced 
 consonant. 
 
 Assimilation and Simplification of a double consonant 
 along with a possible loss of an s. 
 
 § 89.— Examples. 
 
 (a) aspiro from arfspiro, *a5spiro (§§ 68, 56 (2)). 
 
 asto from arfsto, *a5sto {§§ 68, 56 (2)). 
 
 corculum, from *cor(iculom, *corcculom (§§ 76, 56 (3)). 
 
 sarmentmn, from *sarpmeiitom, *sar7nmentom (§§ 70, 
 
 56 (3)). 
 
 arsi, from *arc?si, *arisi (§§ 68, 56 (3)). 
 
 sensi, from *seii^si, *senssi (§§ 68, 56 (3)). 
 
 nox (i.«., noes), from *noc^s (c/. gen. noc^is), *noc5S (§§ 68, 
 
 56 (4). 
 
 testis, from *terstis ('qui tertius stat'), *tessti3. 
 
 tostus, perf. part. pass, of torreo (originally *torseo (§ 74)), 
 from *torstos, *tosstos. 
 
 v> 
 
 * 
 « 
 
 ;^ 
 
 COMBINATIONS OF CONSONANTS 
 
 87 
 
 The two last examples might seem to violate the law 
 laid down in § 74. But for an r to assimilate a following 
 », the latter must first become voiced, which in the case of 
 *terstis and "^torstus was impossible owing to its position 
 before the breathed / (§ 62). Hence the assimilation has 
 been regressive. 
 
 Superstes is a restoration due to analogy ; the phonetic 
 form supestes is found frequently on vulgar inscriptions. 
 
 (6) pone, prep, and adverb, from *posme (c/. superne), 
 ♦posnne, *posne (§§ 69, 56 (3), 86). 
 
 Loss of a consonant together with a possible loss of an s. 
 § 90. A guttural stop is lost between r or I on the one 
 hand, and t, s, w or w on the other, and also between n and 
 
 a dental stop. 
 
 Examples. 
 
 farsi and fartus (farcio) from *farcsi, *farctos. 
 
 sarsi and sartus (sarcio) from *sarcsi, *sarctos. 
 
 falsi and fultus (fulcio) from *fiiksi, *fulctos. 
 
 ultus (ulciscor) from *ulctos. 
 
 tormentum, from *torg''mentom, *torcmentom (§ 82 ; 
 derived from the root of torquQo). 
 
 fulmen, from *ful9'men (fulg-eo). 
 
 quernus, ' oaken,' from *quercnos (qiiercus). 
 
 quindecim, from *quinjw(e)decim, *quincdecim, *quin9'- 
 decim (§§ 15, 82, 62). 
 
 quintus, from ♦quing^tos, quinctos (§ 82). 
 
 The combination -?^c^ was restored by analogy in 
 junctus, unctus, vinctus : quinctus is also found by the 
 side of quintus especially in proper names, Quinctus, 
 Quinctius, Quinctilis. 
 
 § 91. Every labial or guttural stop was lost before s 
 followed by another consonant. When the final consonant 
 
 ■«5 
 
^ 
 
 88 
 
 LATIN PHONETICS 
 
 of the combination was voiced, s was lost in its turn along 
 with compensatory lengthening of the preceding vowel if it 
 was short (§ 86). 
 
 Examples. 
 
 (a) asporto, from *a;7sporto. 
 
 ostendo, from *ojDstendo. 
 
 suscipio, from *suj9Scapio (with umlaut (§ 10, i. c.)). 
 
 The prefixes ops- and sups- are copied from aps- according 
 to the proportional formula ap- {ab-) : aps- = op- {ob-)y sup- 
 {sub) : X. 
 
 sescenti from sexcenti {x = cs). 
 
 Restorations due to analogy : a^stuli (the phonetic 
 form astuli is attested by the grammarian Charisius,^ 1, 
 p. 237, 2 k), dexter, juxta, sextus, etc. These restorations, 
 however, belonged exclusively to the literary language; 
 the people generally knew only dester, justa, sestus, etc., 
 as is shown (i.) by numerous inscriptions and manuscripts 
 written in the common language, and (ii.) by the Romance 
 tongues (c/ Old French, destre, jouste, sistes). 
 
 yote.— It is interesting to find the same simplification of 
 the combinations guttural stop + 5 + breathed consonant into 
 8 + breathed consonant in modern popular French (c/. the 
 common pronunciations esclure, espliguer, estraire). 
 
 (6) amitto, from *apsniitto. 
 
 ebibo, from exbibo. 
 
 sumo, from *su;?s(e)mo (with syncope (§ 15)). 
 
 jumentum, from jouxmentum (a form attested by the 
 most ancient Latin inscription found in 1899 in the Roman 
 Forum ; (m > w (§ 23)). 
 
 sedecim, from *sea;decim. 
 
 tela, from *tea:la (texo). 
 
 1 [Charisius was a grammarian of the fourth century A.D., who 
 transcnbed passages of Palaemon (35-70 A.D.), the author of the Ars 
 Crrammatica.] 
 
 I 
 
 I 
 I 
 
 « 
 
 ^J 
 
 D 
 
 .^ 
 
 \ 
 
 CONBINATIONS OF CONSONANTS 89 
 
 Loss of the Combination -ns- before a Voiced Consonant. 
 
 § 92. The examples of this modification, which is due to 
 the combination of the laws formulated in §§ 86 and 87, 
 are almost exclusively furnished by the prefix trans- when 
 it precedes a voiced consonant, e.g. : — 
 
 traduco, from traTwduco. 
 trameo, from tra?i5meo. 
 trano, from transno. 
 
 The restorations transduco, transmeo, transno, etc., 
 were employed by the side of these phonetic forms. 
 
VVV^ 
 
 90 
 
 LATIN PHONETICS 
 
 APPENDIX. 
 
 THE SYLLABLE. 
 
 Division of Syllables. 
 
 § 93. The combinations of sounds that we call words 
 are divided into as many sections as they contain vowels 
 (or diphthongs). These sections, separated from each 
 other by a complete or partial movement of closing, or 
 simply by a cessation of the vibrations in the glottis, are 
 called syllables. 
 
 In Latin the end of the syllable lay immediately after 
 the vowel (or diphthong) when this was followed by 
 another vowel or single consonant. When the vowel 
 (or diphthong) was followed by two consonants or a 
 double consonant, the consonantal element was as a 
 rule divided between the preceding and following syllable. 
 An exception was made only for the combination stop + 
 trilled or divided consonant, both of which were attached 
 to the following syllable. Lastly, of three consonants the 
 first and the second belonged to the preceding, the third 
 to the following syllable unless the combination was 
 terminated by a stop followed by a trilled or divided 
 consonant. In this case the break came after the first 
 of the three consonants. 
 
 Examples. 
 
 (i.) (a) me-us, qui-es, quo-ad. 
 
 (6) ca-do, pauper, cae-cus, ro-sa, do-mus, si-nus, 
 a-ra, caelum. 
 
 (ii.) (a) ag-raen, pug-na, prop-ter, tec-tura, ip-se, aes-tas, 
 pis-cis, am-bo, om-nis, men-sa, or-do, pul-vis, 
 but 
 qua-drans, a-trox, fe-bris, su-pra, ae-gre, lu-crum, 
 locu-ples. 
 
 We must set aside compounds like abrumpo, sublatus, 
 in which the combination between the stop and 
 
 I 
 
 ^» 
 
 APPENDIX 
 
 91 
 
 the tnlled or divided consonant was much less 
 strong because the prefixes formed for a very 
 long time distinct words {§ 84). Hence these 
 words were divided, ab-rumpo, sub-latus. 
 
 ^nus^'*il-fi^"^''^ ^* bunch), siccus, pos-sum, an- 
 (iii.) abs-temius, cons-picio, ins-tituo, temp-to, sanc-tus, 
 spectrum, plaus-trum, mem-brum, tem-plum. 
 
 . Such is, at least, the method followed by the best 
 inscriptions and manuscripts, which reflects, no doubt 
 faithfully, the phonetic division. The principle laid down 
 by the Koman grammarians from the fifth century of our 
 era stipulates, it is true, that all the combinations which 
 can begin a word must be joined to the second syllable 
 (^.^., a-spice, a-mnis, ca-stra). Their remarks, however, are 
 merely subleties of the imagination, devoid of any lingu- 
 istic interest. ^ ^ 
 
 Duration of Syllables. 
 
 § 94. A syllable is short when it contains a short vowel 
 followed by a single consonant-^.^., the first syllable of 
 cado, cdquo {qu being a single sound (§ 39) ) • 
 
 A syllable is long :— 
 
 (i.) When it contains a long vowel or a diphthong 
 whatever the nature of the consonantal element 
 that folio ws-^.^, the first syllable of paene, 
 actus (§ 26), faustus. 
 
 (ii.) When it contains a short vowel, followed by a 
 double consonant, or a combination of con- 
 sonants — e.^., the first syllable of messis, cf. 
 m6to (§83); sella, cf, s6deo (§71); neptis, 
 grand-daughter' or 'niece,' cf. ngpos (§33)- 
 tango, c/. tetigi from ^etagi (§ 10, i., c). An 
 exception is formed by the syllables containing 
 a short vowel followed by the combination 
 stop -f trilled or divided consonant, which with 
 the early Latin dramatists are always short, 
 
I 
 
 - -^ 4.< 
 
 92 
 
 LATIN PHONETICS 
 
 e.g., patrgm, l5ctiples. The poets of the classical 
 period, it is true, reckon them sometimes short, 
 sometimes long e.g.y Ovid, Met., xiii. 607 : — 
 
 Et prirao similis volucri, mox vera volucrls. 
 
 The lengthening in this case, however, is known to be 
 due to erudite imitation of Greek prosody. 
 
 [In repeating a word the Roman poets constantly vary 
 the quantity in order that the ictus may fall differently on 
 it. Cf. Vergil, Aen., 2, 663 : gnatum ante ora patris, 
 patrem qui obtruncat ad aras ; Hor., Oc?., i. 32, 11 : et 
 Lycum nigris oculis nTgroque. Cf. Monro on lAicr.^ 4, 
 1259 : crassane conveniat liquidis et liquida crassis.] 
 
 § 95. According to the Roman grammarians, the 
 syllables which are long in spite of the short quantity 
 of the vowel they contain, are long positione or positu, 
 Cf. Quintilian, Instit. Orat., 9, 4, 86 : certe in dimen- 
 sione pedum syllaba quae est brevis insequente vel brevi 
 alia, quae tamen duas priores consonantes habeat, fit 
 longa, ut : agrestem tenui musam ... a brevis, gres 
 brevis, faciet tamen longam priorem. dat igitur illi aliquid 
 ex suo tempore. Quo modo, nisi habet plus quam quae 
 brevissima, qualis ipsa esset detractis consonantibus ? 
 nunc unum tempus accommodat priori et unum accipit 
 a sequente ; ita duae natura breves positione sunt 
 temporum quattuor; cf. also Aulus Gellius, Noct. 
 Attic, iv. 17, 8, who says, in reference to suhicit, that 
 the i vim consonantis capit et idcirco ea syllaba produc- 
 tius latiusque paulo pronuntiata priorem syllabam brevem 
 esse non patitur, sed reddit eam positu longam. The 
 terms positione or positu, translated from the Greek, like 
 the majority of the technical terms of Latin grammar, 
 mean properly ' by convention.' The Greek scholars 
 of rhythm, in fact, imagined that the syllables which 
 enclosed a long vowel were naturally long by them- 
 selves ; while the syllables which enclosed a short vowel 
 followed by a double consonant, or a combination of con- 
 sonants, were long through an arbitrary whim — a free and 
 
 '*) 
 
 ri 
 
 o 
 
 APPENDIX 
 
 93 
 
 voluntary arrangement. But the Romans, who were indif- 
 ferent to the question discussed by the Greek philosophers, 
 as to whether names have been attached to things through 
 natural inspiration or artificial convention, soon misunder- 
 stood the meaning of positione or positu, and used these 
 terms in the sense 'by position,' i.e., by the position of the 
 vowel before a repeated consonant or a combination of 
 consonants. 
 
 The explanation of the lengthening ' by position ' lies in 
 the phonetic division of the syllables. A consonantal ele- 
 ment shared between two syllables lengthens the preceding 
 syllable, an consonantal element not shared between them 
 leaves it short. It is not the consonants which add length 
 but the pause which separates them. Note also that it is 
 not the vowel but the syllable which is long by position 
 
 [SPECIMENS OF OLD LATIN. 
 
 L 
 
 From the Laws of the Twelve Tables, 451 B.C. (see 
 Wordsworth, Fragments and Specimens of Early 
 Latin, p. 254). ^ 
 
 (1) Si in ius vocat, ito. Ni it, antestamino ; igitur em 
 capito. (2) bi calvitur pedemve struit, manum endo 
 lacito. (3) bi morbus aevitasve vitium escit (qui in 
 lus vocabit) lumentum dato ; si nolet arceram ne sternito 
 I.e., (1) If he (the accuser) summons (him, the accused) 
 before the law let him go. Unless he goes let him 
 (the accuser) summon him (a bystander) to witness* 
 thereupon let him (the accuser) take him (arrest him)! 
 (2) It he (the defendant) tries to shirk or runs away let 
 him (the accuser) lay his hand upon him. (3) If illness 
 or old age shall be his weakness, let him who summons 
 him grant him a vehicle ; if he is not willing let the 
 accuser not prepare an ambulance. 
 
 h% 
 
4) 
 
 
 ' W "J •«*■ 
 
 
 94 
 
 LATIN PHONETICS 
 
 11. 
 
 APPENDIX 
 
 95 
 
 Sepulchral Inscription in Saturnian Verse. 
 
 L. Cornelias L. f(ilius) Scipio, Consul 259, Censor 258 ; 
 C* 1' Jj.f 1. 32. 
 
 hone oino ploirume cosentiont Romai 
 duonoro optumo fuise uiro 
 Luciom Scipione. filios Barbati 
 consol censor aidilis hie fuet apud uos 
 hec eepit Corsica Aleriaque urbe 
 dedet Tempestatebus aide meretod. 
 
 hunc unum plunmi consentiunt Romae 
 honorum optimum fuisse virum 
 Lucium Scipionem. Jilius Barbati 
 consul f censor, aedilis hie fuit apud vos ; 
 hie eepit Corsicam Aleriamque urbem, 
 dedit Tempestatihus aedem merito. 
 
 III. 
 
 Senatuseonsultum de Bacchanalibus of the year 186 ; 
 a L L. i., 196. 
 
 Q. Marcius L. f. S. Postumius L. f. cos. senatum con- 
 soluerunt N. Octob. apud aedem Duelonai. Sc. arf. M. 
 Claudi M. f. L. Valeri P. f. Q. Minuci C. f. de Bacana- 
 libus quei foideratei esent ita exdeieendum censuere. 
 neiquis eorum Bacanal habuise uelet. seiques esent quei 
 sibei deicerent necesus ese Bacanal habere eeis utei ad 
 pr. urbanum Romam uenirent deque eeis rebus ubei 
 eorum uerba audita esent utei senatus noster decerneret 
 dum ne minus senatoribus C adesent quom ea res coso- 
 leretur. Bacas uir nequis adiese uelet ceiuis Romanus 
 neue nominus Latini neue socium quisquam nisei pr. 
 urbanum adiesent isque de senatuos sententiad dum ne 
 minus senatoribus C adesent quom ea res cosoleretur 
 iousiset. censuere. sacerdos nequis uir eset. magister 
 neque uir neque mulier quisquam eset. neue pecuniam 
 
 "il 
 
 a' 
 
 Of 
 
 
 i 
 
 <4 
 
 quisquam eorum comoinem habuise uelet neue magistra- 
 tum neue pro magistratud neque uirum neque mulierem 
 quisquam fecise uelet. neue post hac inter sed coniou- 
 rase neue comouise neue conspondise neue eonpromesise 
 uelet neue quisquam fidem inter sed dedise uelet. sacra 
 m oquoltod ne quisquam fecise uelet neue in poplicod 
 neue in preiuatod neue exstrad urbera sacra quisquam 
 fecise uelet nisei pr. urbanum adieset isque de senatuos 
 sententiad dum ne minus senatoribus C adesent quom 
 ea res cosolereretur iousiset. censuere. homines pious 
 y omuorsei uirei atque mulieres sacra ne quisquam 
 fecise uelet neue inter ibei uirei pious duobus mulieri- 
 bus pious tribus arfuise uelent nisei de pr. urbani sena- 
 tuosque sententiad utei suprad scriptum est. haiee utei 
 in eouentionid exdeicatis ne minus trinum noundinum 
 senatuosque sententiam utei scientes esetis. eorum 
 sententia ita fuit sei ques esent quei aruorsum ead fe- 
 cisent quam suprad scriptum est eeis rem caputalem 
 faciendam censuere. atque utei hoce in tabolam ahenam 
 inceideretis ita senatus aiquom censuit uteique eam figier 
 ioubeatis ubei faciluraed gnoscier potisit atque utei ea 
 Bacanalia sei qua sunt exstrad quam sei quid ibei saeri 
 est ita utei suprad scriptum est in diebus X quibus 
 uobeis tabelai datai erunt faciatis utei dismota sient. 
 
 Q. Marcius L. f(ilius), S(purius) Postumius L. f(ilius) 
 cons(ules) senatum consuluerutit Nfofiis) Octob(ribus) apud 
 aedem Bellonae. Sc(ribendo) adf(uerunt) M. Claudi(us) M. 
 /(ilius), L. Valeri(us)F./(iliusX Q. Miniici(us) C.J(ilius). 
 de Bacchanalibus qui foederatl essent ita edicendum cen- 
 suere. nequis eorum Bacchanal habuissevellet. siqui essent 
 qui sibi dicerent necesse esse Bacchaiml habere, ei uti ad 
 pr(aetorem) urbanum Romam venii-ent, deque eius rebus, 
 ubi eorum verba audita essent, uti senatus noster decerneret 
 dum ne minus senatoribus C adesseni cum ea res consulerefur. 
 Bacchas vir nequis adiisse vellet civis Romanus, neve nomifiis 
 Latini, neve socioi^um quisquam, nisi pr(aetorem) urbanum 
 adiissent, isque de senatus sententia, dum ne minus senatoribus 
 C adessent cum ea res consuleretur, jussisset. censuere. 
 sacerdos iiequis vir esset. magister neque vir neque mulier 
 
i 
 
 96 
 
 LATIN PHONETICS 
 
 quaequam esset. neve pecuniam quisquam eorum communem 
 
 habuisse vellet, fieve magistratum^ neve pro magistratu neque 
 
 virum neque mulierem quisquam Jecisse vellet. ?ieve posthac 
 
 inter se conjurasse neve convovisse neve conspondisse neve 
 
 compromisisse vellet neve quisquam fidem inter se dedisse 
 
 vellet. sacra in occulta ne quisquam fecisse vellet^ neve in 
 
 publico neve in privato, neve exti-a urbem sacra quisqtiam 
 
 fecisse vellet, nisi pr(aetorem) urbanum adiisset, isque de 
 
 senatus sentential dum 7ie minus senatoribus C adessent cum 
 
 ea res consuleretur, jussisset. Censuere. homines plus V uni- 
 
 versij viri atque mulieres, sacra ne quisquam fecisse vellet, 
 
 neve interibi viri plus duobus mulieri bus plus tribus adfuisse 
 
 vellent, nisi de pr(aetoris) urbani senatusque sententia, uti 
 
 supra scriptum est. kaec uti in contione edlcatis ne minus 
 
 trinum nundinum, senatusque sententiam uti scientes essetis. 
 
 eorum sententia itafuit: siqui essent qui adversum eafecis- 
 
 sent, quam supra scriptum est, eis rem capitalem faciendum 
 
 censuere. atque uti hoc in tabulam ahenam incideretis, ita 
 
 senatus aequum censuit, utique eam figi jubeatis, ubifacillime 
 
 nosci possit, atque uti ea Bacchanalia, siqua sunt, extra 
 
 quam siquid ibi sacri est, in diebus X quihus vobis tabellae 
 
 datae erunt faciatis uti dimota sint. 
 
 IV. 
 
 A Christian Epitaph to a Husband. 
 
 (in Museo Laterensi : de Rossi, n. 62) Fourth Century a.d. 
 
 Depossio luniani pri idus Apriles 
 
 Marcellino et Probino conss. 
 
 qui bixit annis XL. in pace decissit et 
 
 amator pauperorum vixit 
 
 cum Birginia annis xv. benemerenti Birginia sua Bictora 
 
 Benemerenti fecit amatrix pauperorum et operaria. 
 
 i.e. 
 
 Depositio Juniani, pridie idus Apriles, Marcellino et 
 Probino consulibus, qui vixit annos XL. In pace decessit 
 et amator paupermn vixit cuni Virginia annos xv. 
 Bene merenti Virginia sua Victoria bene merenti fecit 
 ainatrix pauperorum et operaria.'^ 
 
 (, 
 
 J\ 
 
 INDEX. 
 
 \_The references are to the sections,], 
 
 a, 67 
 
 ab, 67, 91 
 abdoucit, 22 
 abduco, 62 
 abeis, 30 
 abeo, 30 
 abicio, 15 
 abies, 12 
 abietis, 12 
 abigo, 10 (1) c 
 abis, 30 
 abjectus, 13 (1) 
 abnego, 69 
 abnuo, 69 
 abs, 67 
 
 abstineo, 10 (1) h 
 abstuli, 91 
 
 ac, 31 (1), 82 
 accedo, 56 (2) 
 accipio, 10 (3) 
 accresco, 56 (2) 
 accus(8)o, 14, 22 
 actito, 26 
 actus, 26, 62 (1) 
 
 ad, 38, 56 (2) 
 adactus, 26 
 adbibo, 76 
 ades, 56 (4) 
 adfero, 67, 71 
 adfigo, 67 
 adgero, 76 
 adhibeo, 10 (3) 
 adicio, 15 
 adlatus, 71 
 adloquor, 71 
 
 admoveo, 70 
 adnoto, 69 
 adopto, 14 
 adsequor, 68 
 adspiro, 56 (2), 89 
 adsto, 89 
 adsum, 68 
 adtenuo, 84 
 adtuli, 84 
 aedem, 20, 46 (2) 
 aedes, 20 
 
 aedificium, 10 (1) c 
 aediis, 20 
 aere, 38 
 aeri, 38 
 aeris, 41 
 aes, 38, 41 
 Aeserninu(s), 42 
 affero, 67, 71, 84 
 affigo, 67 
 agceps, 6 (iii.) 
 agcora, 6 (iii.) 
 agellus, 73 
 ager, 73 
 agger 0, 76 
 agguliis, 6 (iii.) 
 agilis, 29 {a) 
 
 ago,10(l),26,29(.-), 
 
 62 (1) 
 ahenum, 35 
 aide, 20, 46 (2) 
 aides, 20 
 Ai(i)ax, 48 
 ai(i)o, 48 
 aiiunt, 48 
 airid, 38 
 aliud, 38 (1) 
 
 alius, 38 (1) 
 allatus, 71 
 alloquor, 71 
 alvearia, 32 
 ama, 31 (3) 
 amare, 41, 42, 56 (4) 
 amaris, 42 
 amas, 56 (4) 
 ambages, 26 
 amitto, 91 
 ammoveo, 70 
 aranego, 69 
 aranuo, 69 
 anas, 44 
 anceps, 6 (iii.) 
 anchora, 35 
 ancora, 6 (iii.), 
 Ancu(s), 42 
 angulus, 6 (iii.) 
 animal, 31 (2) 
 animalis, 31 (2) 
 annecto, 69 
 annitor, 69 
 annoto, 69 
 annus, 13 (1) 
 anser, 44 
 ap-, 91 
 apere, 32 
 aperio, 62 
 
 35 
 
 apio, 
 
 21 
 
 apiscor, 21 
 Apollinis, 29 (6) 
 Apolones, 29 (6) 
 appeto, 1 
 aps-, 91 
 
 aptus, 13 (1), 21 
 aquai, 30 
 
 '^1 
 
 ^J^BS^^"'^^!^,*^^'^ 
 
 msm 
 
98 
 
 arbor, 48 
 
 arbos, 43 
 
 arceo, 13 (1) 
 
 arcubas, 10 (3) 
 
 arcus, 10(3) 
 
 ardeo, 56 (3), 74, 89 
 
 ardere, 15 {a) 
 
 arenam, 41 
 
 aridus, 15 (a) 
 
 aries, 12 
 
 arietis, 12 
 
 arsi, 56 (3), 74, 89 
 
 anindo, 44 
 
 asa, 41 
 
 ascendo, 13 (1) 
 
 asenani, 41 
 
 asiniis, 41 
 
 aspiro, 56 (2), 89 
 
 asporto, 91 
 
 assecla, 82 
 
 assequor, 68, 82 
 
 assigno, 68 
 
 assisto, 68 
 
 assum, 68 
 
 asto, 89 
 
 astuli, 91 
 
 atfeio, 67 
 
 atfigo, 67 
 
 atque, 31 (1), 82 
 
 atapiro, 56 (2) 
 
 attenuo, 8 1 
 
 attingo, 14, 56 (2) 
 
 attraho, 56 (2) 
 
 attuli, 84 
 
 auceps, 10, 15 (^), 
 
 29(c) 
 aucupis, 10 
 audiit, 27 
 and ire, 41 
 aiidivit, 27 
 aufero, 91 
 aula, 59 (2) 
 aulai, 30 
 aulla, 59 (2) 
 aurai, 30 
 aiirea, 32 
 aurifex, 10, 29 (c) 
 aurufex, 19 
 aurum, 29 (c) 
 
 LATIN PHONETICS 
 
 auspex, 1, 10 (1) 6, 
 
 86 
 auspicis, 1, 10 (1) 6, 
 
 86 
 auturapnus, 85 
 auxiliaris, 45 (1) 
 ave, 31 (3) 
 avere, 31 (3) 
 avis, 29 (r) 
 
 B 
 
 bac(c)a, 55 
 Bacchus, 35 
 barba, 13 (1) 
 basis, 41 
 benes, 31 (3), 36 
 beni, 36 
 bictor, 52 
 biduom, 29 
 biduum, 29 
 biennium, 13 (1) 
 biforis, 14 
 biginti, 52 
 bixi, 36 
 bonorum, 46 (2) 
 
 cado, 10 (i.) r, 26, 83 
 caedo, 14, 59 (2), 60, 
 
 70 
 caelum, 45 (2), 59 (2), 
 
 71 
 caeraentum, 60, 70 
 Caepio, 35 
 caeruleus, 45 (2) 
 calcar, 31 (2) 
 calcaris, 31 (2) 
 caldus, 15 {a) 
 calicis, 29 (a) (1) 
 calidus, 15 {a) 
 calix, 29 (a) (1) 
 canalia, 56 (1) 
 canna, 56 (1) 
 cano, 10 (1) c 
 cantabam, 31 (2) 
 
 cantabas, 31 (2) 
 cantai, 51 
 cape, 31 (1) 
 capio, 10 (3), 29 (c) 
 capitis, 10 (1) c 
 caput, 10 (1) e 
 carpo, 13 (1) 
 cas(s)u8, 26, 41, 58, 
 
 83 
 cate, 31 (8) 
 catus, 33 (x.) 
 caus(s)a, 14, 22, 41, 
 
 58 
 -ce, 76 
 
 cecTdi, 10 (1) c 
 cecidi, 14 
 cecini, 10 (1) c 
 cena, 6 (i.) 
 cera, 20 
 Cereres, 29 (b) 
 Cereris, 29 (b) 
 cervical, 45 (1) 
 cesor, 87 
 Cetegus, 35 
 Cethegus, 35 
 ceu, 22 
 chenturio, 35 
 chommoda, 35 
 chorona, 35 
 chorus, 6 (iy.) 4 
 cineris, 11 
 cinis, 11 
 cisium, 41 
 cithara, 6 (iv. ) 4 
 cito, 10(1) a, 31 (3) 
 civis, 6 (i.) 
 clam, 32, 80 
 clanculum, 80 
 clas(8)is, 58 
 claudo, 14, 63, 85 
 clausi, 63 
 claustrum, 85 
 coalesce, 32 
 coctus, 82 
 codiux, 53 
 coegi, 32 
 coepi, 21, 32 
 cocrceo, 13 (1) 
 coetus, 32 
 
 I 
 
 t 
 
 c 
 
 I 
 
 cogo, 32 
 coitus, 32 
 colligo, 1, 10 (1) b 
 colloquium, 72 
 collum, 66, 74 
 columen, 18 
 com-, 80 
 comedo, 1, 14 
 comes, 80 
 
 comminuo, 10 (1) « 
 commoda, 35 
 commoinis, 21 
 communis, 21 
 como, 32, 85 
 comoinem, 21 
 compitum, 10 (1) & 
 complodo, 24 
 compono, 80 
 compsi, 85 
 comptus, 85 
 con-, 25 
 concenatio, 80 
 concino, 10 (1) c 
 concrepuit, 25 
 concussi, 68 
 concutio, 15 (6), 68 
 con feci, 15 
 con fecit, 25 
 confectum, 15 
 confectus, 26 
 conficio, 10 (l)c, 15, 
 
 33 
 
 confiteor, 10 jl) c 
 
 conflovout, 47 
 
 confringo, 16 
 
 conicio, 15 
 
 coujeci, 15 
 
 conjee turn, 15 
 
 conjicio, 15 
 co(n)jux, 53 
 conloquium, 72 
 compono, 80 
 conposuit, 25 
 conquinisco, 76 
 conscendo, 13 (1) 
 consentio, 13 (3), 29 
 
 (d) 
 co(n)8entiont 29 (<?) 
 consentiunt, 29 {d) 
 
 INDEX 
 
 conservos, 13 (3) 
 consilium, 87 
 conspicio, 56 (2) 
 constiti, 1 
 consuevit, 25 
 coHsul, 87 
 consularis, 45 (1) 
 consules, 25 
 contempno, 85 
 contexo, 80 
 conticesco, 10 (1) c 
 continues, 10 (1) b 
 contio, 32, 50 
 contubernalis, 10 (3) 
 convoco, 14 
 copia, 32 
 copula, 32 
 coquo, 82 
 coram, 32 
 
 corculum, 56 (3), 89 
 cordis, 56 (3) 
 corniger, 10 (1) « 
 cornu, 10 (1) c 
 corona, 35 
 corporis, 14 
 corpus, 14 
 corripio, 72 
 corrumpo, 72 
 corumpo, 72 
 coruptus, 72 
 cos, 33 (x.) 
 cosol, 87 
 coventio, 32, 50 
 crebesco, 45 (4) 
 crebresco, 45, (4) 
 crebrui, 45 (4) 
 crebui, 45 (4) 
 cubital, 45 (1) 
 culm en, 18 
 cumulare, 85 
 1 cup(p)a, 55 
 ' currus, 30, 56 (1) 
 curulis 56 (1) 
 curvos, 13 (3) 
 
 / 
 
 dabi(s), 42 
 dacrima, 37 
 
 99 
 
 dare, 11, 38 (2) b 
 dato, 38 (2) b 
 datod, 38 (2) b 
 dator, 43 
 datus, 10(l)c,33(x.) 
 
 dautia, 37 
 deamo, 32 
 debere, 36 
 decem, 50, 80 
 decet, 16 
 decido, 10(l)c 
 decoris, 14 
 decus, 14 
 dedccus, 14 
 dederont, 29 {d) 
 dederunt, 29 {d) 
 dedi, 10(i.)6 
 dedicait, 51 
 dedisco, 13 (3) 
 deesse, 32 
 defero, 40 
 degi, 32 
 dego, 32 
 dehinc, 32 
 deicerent, 19 
 deico, 19 
 deinde, 32 
 delere, 41 
 delictum, 82 
 deligo, 10 (1) b 
 delinquo, 82 
 -dem, 80, 86 
 demo, 32, 85 
 dempsi, 85 
 demptus, 85 
 dens, 33 (ii.) 
 deorsum, 50 
 descendo, 13 (1) 
 deses. 29 (&) 
 desidis, 29 {h) 
 desilio 10 (2) 
 despicio, 56 (2) 
 desse, 32 
 dester, 91 
 destiti, 10 (1) b 
 desuper, 41 
 detrectro, 13 (1) 
 dexter, 91 
 dextrorsum, 49 
 
n 
 
 100 
 
 LATIN PHONETICS 
 
 die, 31 (1) 
 
 dico, 19, 85 (2) 
 
 dicsi, 85 (2) 
 
 dictus, 85 (2) 
 
 diduco, 86 
 
 die, 38 (2) a 
 
 diei, 27 
 
 dies, 27, 38 (2) a 
 
 diifeideus, 19 
 
 diffeido, 19 
 
 differo, 79 
 
 difticilis, 10(l)(r, 79 
 
 diffido, 19, 79 
 
 ditiiteor, 10(l)c 
 
 diffundo, 79 
 
 digero, 86 
 
 diguus, 6 (iii.)» 16» 
 
 69 
 dilanio, 86 
 dimico, 10 (1) a 
 dimidius, 10 (1) h 
 diraitto, 86 
 dingua, 37 
 dinosco, 25 (1), 86 
 diribeo, 44 
 dirimo, 41 
 dis, 49 
 dis, 56 (2) 
 discedo, 86 
 discerpo, 13 (1) 
 discindo, 56 (2) 
 disciplina, 15 {a) 
 discipulus, 15 (a) 
 disco, 13 (3) 
 disertus, 56 (1) 
 dispicio, 56 (2) 
 displodo, 24 
 dispone, 86 
 dispute, 14 
 disque tulisseiit, 84 
 disscindo, 56 (2) 
 disseco, 1 
 dissentio, 13 (3) 
 dissimilis, 56 (2) 
 dissocio, 86 
 disspicio, 56 (2) 
 disto, 56 (2) 
 distorqueo, 86 
 dite, 49 
 
 diti, 49 
 ditior, 49 
 ditis, 49 
 ditissimus, 49 
 diu, 80 
 
 dives, 29 (6), 49 
 divide, 83 
 
 divis(s)io, 41, 58, 83 
 divitior, 49 
 divitis, 29 (6), 49 
 divitissimus, 49 
 dixi, 85 (2) 
 doce, 31 (3) 
 documentum, 10 (3) 
 donis, 30 
 denem, 30 
 donum, 30, 33 
 due 31 (I) 
 duco, 22 
 duo, 47 
 duvo, 47 
 dvoiioro(m), 46 (2) 
 
 ! e-, 67 
 I earn, 80 
 
 earadem, 80 
 
 eandem, 80 
 I earum, 80 
 j earum dem, 80 
 j earundem, 80 
 I ebibo, 91 (6) 
 ' ec-, 67 
 
 ecfari, 67 
 
 ecferet, 67 
 
 editus, 10(l)c 
 
 edo, 1, 14, 33(ii.), 83 
 
 educe, 25 (1) 
 edus, 20 
 
 effectus, 13(i.), 26 
 effero, 67 
 efferus, 11 
 effunde, 13 (3) 
 egenus, 86 
 egestas, 31 (3), 86 
 eiius, 48 
 ejus, 86 
 
 ejusdem, 86 
 
 emi, 33 (iv.) 
 
 eraico, 10 (1) a 
 
 emo, 33 (iv.) 
 
 endo, 13 (2) 
 
 endostruos, 13 (2) 
 
 ensiuni 29 (a) 
 
 ensis 29 («) 
 
 eo, 31 (3) 
 
 eorum, 80 
 
 eerumdem, 80 
 
 eerundem, 80 
 
 eque, 33 (i.) 
 
 equom, 39, 46 (4) 
 
 equos, 29, 33 (1) 
 
 equus, 29 
 
 erigo, \0 {\)h 
 
 erit, 41 
 
 erus, 44 
 
 es, 56 (4) 
 
 esse, 38 (2) &, 41, 56 
 (4), 58, 65 
 
 est, 33 (ii.), 41 
 
 essus, 83 
 
 etiam nunc, 46 (i) 
 
 eum, 80 
 eumdem, 80 
 eundeni, 80 
 evideus, 10 (1) a 
 ex-, 56 (3), 67 
 excerpo, 13 (1) 
 excite, 10 (1) a 
 exemplum, 85 (2) 
 exercee, 13 (1) 
 exilium, 10 (2) 
 eximo, 85 (2) 
 explode, 24 
 exsatie, 56 (3) 
 ex(3)culpo, 56 (3) 
 exsilium, 10 (2) 
 exselve, 56 (3) 
 exsomnis, 56 (3) 
 ex(s)pelio, 56 (3) 
 ex(s)truo, 56 (3) 
 exsulans, 10 (2) 
 exsulto, 17, 33 
 extinguo, 82 
 extinxi, 82 
 exulans, 10 (2) 
 
 >! 
 
 \" 
 
 r\ 
 
 I 
 
 faba, 36 
 
 Fabaris, 45 (5) 
 
 fac, 31 (1) 
 
 facetus, 10 (1) c 
 
 facilis, 14 
 
 facie, 10 (1) c, 26,29 
 
 (0, 33, 33 (ix.) 
 facis, 29 
 factito, 26 
 factus, 13 (1), 26 
 failla, 51 
 Falerii, 11 
 Falisci, 11 
 fallo, 13 (1), 40 
 familia, 10 (2) 
 famulus, 10 (2) 
 farcie, 74, 90 
 Farfarus, 45 (5) 
 farsi, 74, 90 
 fartus, 90 . 
 fateer, 10 (1) c 
 fa villa, 61 
 fax, 29 
 feci, 33 (ix.) 
 fefelli, 13(1), 40 
 feido, 33 (iii. ) 
 fel, 56 (4) 
 fellis, 56 (4) 
 fer, 31 (1) 
 fere, 31 (3) 
 ferentarium, 31 (3) 
 feriae, 41 
 ferimus, 10 (3) 
 fere, 15 (6) 
 ferre, 17, 65, 74 
 ferrus, 11 
 fesiis, 41 
 festus, 41 
 fhefhaked, 6 (iii.) 
 fides, 33 (iii.) 
 fido, 33 (iii.) 
 filios, 29 id) 
 filius, 29 {d) 
 fimus, 27 
 finio, 27 
 finire, 27 
 
 INDEX 
 
 fie, 27 
 fis, 27 
 
 firmus, 13 (3) 
 flagrare, 45 (3) 
 Flaurus, 24 
 flebam, 27 
 flee, 27 
 Florus, 24 
 fluere, 47 
 
 flumen,10(l)6, 29(^) 
 fluminis, 10 (1) &, 
 
 29 (&) 
 fluvere, 47 
 fluvius, 47 
 fodi, 33 (vii.) 
 fedio, 26, 33 (vii.) 
 foederum, 41 
 feedesum, 41 
 feedus, 33 (iii.) 
 foidere, 33 (iii.) 
 foidos, 33 ,iii. ) 
 feres, 14 
 forfex, 29 (6) 
 forficis, 29 {b) 
 foris, 14 
 fossus, 26 
 fovea, 50 
 fragrare, 45 (3) 
 frigora, 28 
 frigus, 28 
 fruontur, 13 (2) 
 fruor, 47 
 fruuntur, 13 (2) 
 fruver, 47 
 fulcio, 90 
 fulcrum, 45 (1) 
 fulgeo, 90 
 fulsi, 90 
 fulmen, 90 
 fultus, 90 
 fundo, 13 (3) 
 Furii, 41 
 Fusii, 41 
 
 gaesum, 41 
 gallina, 56 (1) 
 
 101 
 
 gallus, 56 (1) 
 genera, 28 
 generis, 29 (6) 
 genetricera, 33 (\iii.) 
 genetrix, 33 (viii.) 
 geniter, 33 (viii.) 
 genitorem, 33 (viii.) 
 genui, 33 (ii.) 
 genus, 29 (6) 
 gere, 58 
 gessi, 58 
 gigne, 33 (ii.) 
 Gianuarius, 53 
 Giove, 53 
 glacies, 27 
 glaciei, 27 
 glube, 60, 70 
 gluma, 60, 70 
 Gracchus, 35 
 Graccus, 35 
 gradior, 71 
 grallae, 71 
 gruis, 27 
 grus, 27 
 
 H 
 
 habeo, 10 (3), 14 
 haedus, 20 
 banc, 76 
 hance, 76 
 harundo, 44 
 haruspex, 29 (6) 
 haruspicis, 29 {b) 
 baud, 38 
 haurio, 41 
 haustum, 41 
 have, 31 (3) 
 havere, 31 (3) 
 heme, 32 
 heri, 20, 41 
 herus, 44 
 hesternus, 41 
 hie, 56 (4), 76 
 hiemis, 29 (b) 
 hieras, 29 (&) 
 hinsidias, 35, 44 
 hirpex, 44 
 
 G* 
 
102 
 
 hoc, 56 (4), 76 
 holus, 17, 44 
 homo, 32 
 honce, 76 
 honestus, 33 (vi.) 
 honor, 33 (vi.), 43 
 hoDorem, 33 (vi.) 
 honos, 43 
 humerus, 44 
 hunc, 76 
 
 I 
 
 iccirco, 76 
 
 id, 76 
 
 idcirco, 76 
 
 idem, 25(1), 63, 86 
 
 iferos, 87 
 
 ilico, 10 (1) d 
 
 illacrimor, 14 
 
 illatus, 72 
 
 ille, 38 (1) 
 
 illi, 27 
 
 illido, 14 
 
 illis, 30 
 
 illius, 27 
 
 illud, 38 (1) 
 
 irabellis, 80 
 
 imberbis, 13 (1) 
 
 imbuo, 80 
 
 immerito, 81 
 
 iraminuo, 10 (1) a 
 
 iramitis, 81 
 
 immolo, 81 
 
 immuto, 81 
 
 impudicus, 14 
 
 in-, 25, 80 
 
 iDceideretis, 19 
 
 inceido, 19 
 
 incido, 10 (l)c 
 
 incido, 19 
 
 incohare, 35 
 
 incomparabilis, 36 
 
 incomparavilis, 36 
 
 indoctus, 25 
 
 induro, 80 
 
 industrius, 13 (2) 
 
 ineo, 80 
 
 LATIN PHONETICS 
 
 ineptu3, 13 (1) 
 infantem, 87 
 infelix, 25 
 infero, 72 
 iuferus, 15 {a), 33 
 inficetus, 10 (1) c 
 infirmus, 13 (3) 
 infra, 15 (a), 33 
 ingenuos, 80 
 inhumanus, 25 
 immolo, 81 
 innocuous, 56 (1) 
 innoxius, 56 (1) 
 inopia, 32 
 inquietus, 80 
 inrevocabilis, 72 
 insanus, 25 
 insidia8, 35, 44 
 insimul, 87 
 insulsus, 14 
 intactus, 80 
 intibum, 10 (3) 
 introrsura, 49 
 intubum, 10 (3) 
 inutilis, 80 
 invicem, 10 (1) a 
 Iphianassai, 30 
 ipsi, 27 
 ipsius, 27 
 irpex, 44 
 irrevocabilis, 72 
 irrigo, 10 (1) a 
 irriguos, 10 (1) a 
 irritus, 10 (1) c 
 is, 86 
 
 ita, 10 (1) c, 28 
 itidem, 10 (1) c 
 
 ; Julia, 53 
 jumentum, 91 (6) 
 junctus, 90 
 jungo, 90 
 junior, 15 (6), 32 
 Jup(p)iter, 53, 55 
 juris, 41 
 jus, 41 
 justa, 91 {a) 
 Justus, 53 
 juvenis, 32 
 juventa, 47 
 juxta, 91 {a) 
 
 K 
 
 K = Kaeso, 6 (i.) 
 
 K or Kal = Kalendae, 
 
 6 (i.) 
 KA = capitalis, 6 (i.) 
 
 kaput, 6 (i.) 
 karus, 6 (i.) 
 Kartago, 35 
 KK = castrorum, 6(i.) 
 
 jacio, 15 
 jactus, 13 (1) 
 Januarius, 53 
 jouxmentom, 91 (6) 
 Jove, 53 
 judex, 86 
 judicis, 86 
 
 labor, 43 
 labos, 43 
 lac, 29 
 lachrima, 35 
 lacrima, 35, 37 
 lacrimo, 14 
 lactis, 29 
 laedo, 14 
 lases, 41 
 lasibus, 41 
 latrina, 32, 49 
 lautia,. 37 
 lautus, 37 
 lavacrum, 45 (1) 
 lavatrina, 32, 49 
 lebare, 36 
 lectus, 26 
 legere, 11, 41 
 legerupa, 11 
 legifer, 11 
 
 1 
 
 ♦ 
 
 >j 
 
 jI K )j 
 
 v" 
 
 legimus, 10 (3) 
 lego, 1, 10 (1) 6, 16, 
 
 26 
 Lemuria, 45 (3) 
 lene, 28 
 leniit, 27 
 lenire, 27 
 lenis, 28 
 lenivit, 27 
 levare, 36 
 liberta, 36 
 libertabus, 36 
 libertavus, 36 
 libet, 10 (3) 
 lictor, 31 (2) 
 lictoris, 31 (2) 
 lignum, 16, 69 
 limpha, 6 (iv. ) 4 
 lingere, 37 
 lingua, 37 
 lit(t)era, 55 
 lit(t)us, 55 
 locus, \0{\) d 
 loucos, 23 
 lubet, 10 (3) 
 Luciom, 46 (2) 
 Lucium, 46 (2) 
 lucrum, 45 (1) 
 lucus, 23 
 ludo, 68 (3) 
 lunaris, 45 (1) 
 lusi, 58 (3) 
 
 M 
 
 machina, 6 (iv.) 4 
 Maesius, 20 
 mage, 42 
 magis, 42 
 
 magistratu, 38 (2) a 
 magistratud, 38 (2) a 
 magistratus, 29 (a), 
 
 38 (2) a 
 magnus, 33 (vi.)» 46 
 
 (1) 
 Maia, 48 
 
 Maiia, 48 
 
 maiiorem, 48 
 
 INDEX 
 
 majestas, 33 (vi.) 
 majorem, 33 (vi.) 
 majosem, 33 (vi.) 
 majosibus, 41 
 majus, 33 (vi.) 
 male, 31 (3) 
 manceps, 10 (3) 
 mancipis, 10 (3) 
 mancupis, 10 (3) 
 manibus, 10 (3) 
 manica, 10 (1) e 
 manifestus, 10 (3) 
 manufestus, 10 (3) 
 manus, 10(1) c, 10(3) 
 mare, 28 
 maria, 28 
 Mato, 35 
 maxinius, 10 
 med, 38 
 
 medius, 10 (1) h 
 mel, 29 
 meliorem, 41 
 meliosem, 41 
 meliosibus, 41 
 mellis, 29 
 merito, 38 (2) a 
 raeritod, 38 (2) a 
 meritum, 38 
 mertare, 74 
 mesem, 87 
 Mesius, 20 
 messis, 83 
 raessui, 68 
 messus, 26 
 meto, 26, 68, 83 
 mi, 32 
 
 mico 10 (1) a 
 mihi, 32 
 miles, 42, 56 (4), 68 
 
 (2) 
 milia, 59 (1) 
 militaris, 45 (1) 
 militis, 68 (2) 
 mille, 59 (1) 
 minimu'i, 10 
 minuo, 10 (1) a 
 missus, 58 
 mitte, 31 (1) 
 mitto, 58 
 
 103 
 
 modo, 31 (3) 
 moenia, 21 
 moerus, 21 
 mola, 81 
 mons, 13 (2) 
 mortalis, 45 (1) 
 muc(c)iis, 55 
 mulceo, 74 
 mulsi, 74 
 
 multum ille, 46 (1) 
 munire, 21 
 raurus, 21 
 
 N 
 
 nausea, 41 
 navalis, 45 (1) 
 nee, 31 (1), 82 
 necis, 29 
 nemo, 32, 44 
 nernpe, 76 
 nepos, 33 (viii.) 
 nepotem, 33 (viii.) 
 neptem, 33 (viii.) 
 neplis, 33 (viii.) 
 neque, 31 (1), 82 
 neu, 22, 31 (1) 
 neuter, 22 
 neve, 22, 31 (1) 
 nex, 29 
 nihil, 32, 44 
 nil, 32 
 ninguit, 39 
 nisi, 41 
 
 nivis, 29, 39, 82 
 nix, 29, 39, 82 
 noctis, 29, 89 
 nonus, 50 
 noutrix, 23 
 novem, 50 
 novitas, 10 (1) <i, 12 
 novos, 10 (1) d, 12 
 novus, \Q \\) d 
 nox, 29, 89 
 nubo, 62 (1) 
 nucis, 29 
 numerum, 85 
 numerus, 11 
 
 I 
 
I 
 
 
 104 
 
 nupsi, 62 (1) 
 nutrix, 23 
 mix, 29 
 
 ob-, 70, 91 (a) 
 obdo, 62 (2) 
 obduro, 64 
 obfero, 67 
 obfundo, 67 
 obgero, 76 
 obicio, 15 
 obicit, 15 
 obsideo, 10 (1) 6 
 obtego, 1 
 obtineo, 64, 67 
 obtrecto, 13 0) 
 occa(n)sio, 87 
 occupo, 10 (3) 
 ocquinisco, 76 
 odi, 33 (vii.) 
 odium, 33 (vii.) 
 odor, 37 
 
 oenus, 21 
 
 ofella, 56 (1) 
 
 offa, 56(1) 
 
 oHero, 67 
 
 otlicina, 67 
 
 officio, 10 (1) r 
 
 offundo, 67 
 
 cggero, 76 
 
 oino(m), 21, 46(2) 
 
 oinos, 21 
 
 oisos, 21 
 
 olere, 37 
 
 olla, 59 (2) 
 
 does, 30 
 
 olus, 44 
 
 omitto, 70 
 
 op-, 70,91 (a) 
 
 operio, 62 (2) 
 
 operis, 41 
 
 opfero, 67 
 
 opfundo, 67 
 
 opifex, 67, 86 
 
 opificina, 67 
 
 opificis, 86 
 
 LATIN PHONETICS 
 
 opos, 29 {d) 
 opprimo, 56 (2) 
 oppugno, 56 (2) 
 ops, 91 (ti) 
 optimus, 10, 42 
 optineo, 64 
 opto, 14 
 opturaus, 10 
 opus, 29 (rf), 41 
 Orcivius, 35 
 ordinis, 29 {b) 
 ordo, 29 {b) 
 oris, 32 
 OS, 29, 56 (4) 
 OS, 32 
 
 ossis, 29, 56 (4) 
 ostendo, 91 {a) 
 Otho, 35 
 Oto, 35 
 
 painientum, 51 
 palam, 32 
 Palilia, 45 (2) 
 pango, 10(l)c 
 paor, 51 
 paorem, 51 
 Fapisius, 41 
 parco, 13 (1) 
 paries, 12 
 parietis, 12 
 
 Parilia, 45 (2) 
 pario, 11 
 paro, 14 
 
 passus sum, 83 
 
 patior, 83 
 
 pausa, 41 
 
 pavimentum, 51 
 
 pavor, 51 
 
 -pe, 76 
 
 pedis, 12, 68 (2) 
 
 pelegrinus, 45 (3) 
 
 pellicio, 73 
 
 pello, 17, 74 
 
 pelluceo, 73 
 
 pendo, 33 (i.) 
 
 penna, 69 
 
 peperci, 13 (1) 
 
 peperi, 11 
 pepigi, 10 (1) c 
 per-, 56 (2) 
 perdisco, 13 (3) 
 peregrinus, 45 (3) 
 perfacilis, 14 
 pergo, 56 (2) 
 periclum, 45 (1) 
 perlicio, 73 
 perluceo, 73 
 perrexi, 56 (2) 
 perrumpo, 56 (2) 
 pertinax, 10 (1) 6 
 per video, 10 (1) « 
 pes, 12, 68 (2) 
 pessimus, 56 (3) 
 peto, 1, 10 (1) />, 69 
 philtrum, 6 (iv.), 4 
 pictai, 30 
 picis, 29 
 pietas, 12 
 pijus, 47 
 
 pios, 12 
 
 pius, 12, 27, 47 
 
 pix, 29 
 
 plaudo, 24, 68 (3), 83 
 
 plausi, 68 (3) 
 
 plaus(s)us, 83 
 
 plaustra, 24 
 
 plebi, 64 
 
 plebis, 64 
 
 plebs, 64 
 
 pleps, 64 
 
 plodo, 24 
 
 plostra, 24 
 
 pluit, 47 
 
 pluriraa, 41 
 
 plusima, 41 
 
 pluvia, 47 
 
 pluvit, 47 
 
 po-, 25(1) 
 
 poclura, 45 (1) 
 
 poena, 21 
 
 Poenus, 21 
 
 pollen, 18 
 
 pondo, 33 (i.) 
 
 pone, 89 
 
 pono, 25 (1), 86 
 
 porrigo, 10 (1) 6 
 
 V 
 
 \ 
 
 n 
 
 > 
 
 c; 
 
 { 
 
 positus, 25 (1) 
 possumus, 10 (3) 
 postbabeo, 14 
 potes, 56 (4) 
 praecho, 35 
 praecino, 10 (1) c 
 praeda, 38 (2) rt 
 praesideo, 1, 10 (1) b 
 praestigiae, 45 (4) 
 praestnngo, 45 (4) 
 praidad, 38 (2) a 
 prelum, 86 
 premo, 86 
 pressi, 86 
 primus, 86 
 priscus, 86 
 probai, 51 
 probavi, 51 
 prodigus, 10 (1) c 
 proditus, 10 (1) c 
 profiteer, 10 (1) c 
 prohibeo, 10 (3) 
 promo, 85 
 
 promontorium, 13(2) 
 prompsi, 85 
 promptus, 85 
 promunturium 13 (2) 
 propietas, 45 (4) 
 propius, 45 (4) 
 proprietas, 45 (4) 
 proprius, 45 (4) 
 protector, 26 
 providus 1 ( 1 ) a 
 pudicus, 14 
 pulcer, 6 (iv.), 35 
 pulcher, 6 (iv. ), 35 
 Pulcher, 35 
 pulsus, 17, 74 
 pultare, 74 
 pulveris, 11 
 pulvinar, 45 (1) 
 pulvis, 11, 18 
 Punicus, 21 
 punire, 21 
 punis, 31 (2) 
 punit, 31 (2) 
 puta, 28, 31 (3) 
 putare, 31 (3) 
 puto, 14 
 
 INDEX 
 
 quae, 6 (i.) 
 qomes, 6 (i.) 
 quaerit, 20 
 quaero, 20 
 quairatis, 20 
 quairo, 20 
 quam, 80 
 quamdiu, 80 
 quandiu, 80 
 quantum erat, 46 (1) 
 quasei, 31 (3) 
 
 quasi, 31 (3) 
 quassus, 83 
 
 quatio, 83 
 
 quercus, 90 
 
 quern us, 90 
 
 queror, 41 
 
 questus sum, 41 
 
 qui, 6(i.) 
 
 quibus, 86 
 
 quibusdara, 86 
 
 quic^uara, 76 
 
 quid, 76 
 
 quidam, 86 
 
 quidquam, 76 
 
 Quinctilis, 90 
 
 Quinctius, 90 
 
 Quinctus, 90 
 
 quindeeim, 90 
 
 quintus, MO 
 
 quippe, 76 
 
 quod, 6 (i.) 
 
 quodlibet, 10 (3) 
 
 qura, 6 (i.) 
 
 rado, 59,60,71,85(1) 
 rallum, 59, 71 
 ramentum, 60 
 rancidus, 14 
 rapio, 10 (3) 
 rastrum, 85 (1) 
 ratus, 10 (1) c 
 raucus, 15 (a) 
 ravis, 15 {a) 
 reccidi, 15 {b) 
 
 105 
 
 recido, 15 (6) 
 recipero, 10 (3) 
 recupero, 10 (3) 
 recurvos, 13 (3) 
 reddere, 11 
 reddidi, 10(1)6 
 reducis, 29 {a) 2 
 redux, 29 («) 2 
 refectus, 13 (1) 
 refero, 15 (6) 
 reficio, 40 
 regalis, 45 (1) 
 rego, 10 (1) 6 
 rei, 27 
 
 rem, 33, 46 (4) 
 remex, 29 c 
 rcmus, 29 (c) 
 re pello, 15 (b) 
 reperio, 11 
 reppuli, 15 {b) 
 res, 27, 33 
 resedo, 41 
 restiti, 10 (1) 6 
 reatituo, 10 (1) c 
 reticeo, 10 (1) c 
 retro, 31 (3) 
 retrorsum, 49 
 rettuli, 15 (//) 
 rexi, 62 (1) 
 rigo, 10 (1) rt 
 romphaea, 6 (iv.), 4 
 rosa, 30 
 rosis, 30 
 rufus, 40 
 
 Sabelli, 69 
 Sabini, 69 
 saccus, 56 (1) 
 sacellus, 56 (1) 
 salio, 10 (2) 
 salsus, 14 
 salto, 33 
 Salutes, 29 (6) 
 Salutis, 29 (b) 
 salvom, 46 (2) 
 salvon, 46 (2) (4) 
 Samnium, 69 
 
 1 t 
 
m^^^W^^^^m 
 
 
 * . ^ 
 
 106 
 
 LATIN PHONETICS 
 
 sarcio, 90 
 
 sarmentum, 56 (3), 89 
 sarpere, 56 (3) 
 sarsi, 90 
 sartus, 90 
 aatiilliis, 73 
 satur, 73 
 satus, 83 (ix.) 
 scabellum, 69 
 scamnum, 69 
 scando, 13 (1) 
 scindo, 56 (2) 
 scribo, 62 (1) 
 scriptus, 62 (1) 
 scrofa, 40 
 ae-, 10 (2) 
 secludo, l4 
 seco, 26, 62 (2) 
 sectio, 26 
 secubo, 57, 76 
 sed, 38(1) 
 sed, 14, 76 
 sedecim, 25 (1), 91 b 
 sedeo, 1,10 (1)^26, 
 33 (v.), 37, 71, 83 
 sedes, 33 (v.) 
 seditio, 76 
 sedo, 41 
 sedulo, 10 (2) 
 segmentum, 62 (2) 
 seligo, 10 (1) b 
 sella, 71 
 semen, 33 (ix.) 
 sensi, 56 (3), 89 
 sentio,13(3),56(3),89 
 
 fceorsum, 50 
 separare, 14 
 separo, 14, 57, 76 
 sepero, 14 
 septem, 80 
 septemdecim, 80 
 septendecim, 80 
 septingenti, 16 
 aepulchrum, 6 (iv. ) 
 
 4, 35 
 sepulcrum, 35, 45 (1) 
 sequontur, 13 (2) 
 serra, 56 (1) 
 ^erratus, 56 (1) 
 
 servos, 13 (3) 
 sescenti, 91 
 sessus, 26, 83 
 
 sestus, 91 
 
 sen, 22 
 
 sextus, 91 
 
 si, 41 
 
 sibei, 80 
 
 sibi, 30, 32 
 
 sibilare, 40 
 
 siccus, 56 (3), 76 
 
 Sicilia, 10 (2) 
 
 Siculus, 10 (2) 
 
 sido, 33 (T.) 
 
 sifilare, 40 
 
 similis, 10 (2) 
 
 simulacrum, 45 (1) 
 
 simulo, 10 (2) 
 
 sis ( = si vis), 49 
 
 sitis, 76 
 
 situs, 25 (1) 
 
 soldus, 15 (a) 
 
 solidus, 15 {a) 
 
 solium, 33 (v.), 37 
 
 somnus, 69 
 
 sompnus, 85 
 
 sopor, 69 
 
 specimen, 10 (3) 
 
 specie, 29 (6), 56 (2) 
 
 speni, 31 (2) 
 
 spes, 31 (2) 
 -spicio, 56 (2) 
 
 spiro, 56 (2) 
 sposa, 87 
 statuo, 10 (I) c 
 Stella, 59(1) 
 stelio, 59(1) 
 steti, 1, 10(1)6 
 stilicidium, 59 (1) 
 stilla, 59(1) 
 stipes, 10 (1) b 
 stipitis, 10 (1) b 
 stiria59(l) 
 stlocus, 10 (l)rf 
 sto, 56 (2) 
 stolidus, 18 
 structum, 39 
 struo, 13 (2), 39 
 struvo, 39 
 
 stultus, 18 
 sub-, 91 (a) 
 subdolus, 64 
 subicio, 15 
 subicit, 15 
 subjectus, 13 (1) 
 submitto, 70 
 subrancidu3, 14 
 substituo, 10 (1) c 
 subtilis, 64, 67 
 subula, 27 
 subveuio, 1 
 succido, 76 
 succiduos, 10 (1) c 
 sulphur, 6 (iv.) 4, 35 
 sulpur, 35 
 suramitto, 70 
 summus, 56 (3), 63, 70 
 sumo, 80, 85, 91 (6) 
 sumpsi, 85 
 sumptus, 80, 85 
 sunt, 33 (ii.) 
 sunto, 38 (2) b 
 suntod, 38 (2) b 
 suntus, 80 
 suo, 27 
 sup-, 91 (a) 
 super, 41 
 superue, 89 
 superstes, 89 
 suprrior, 63 
 superus, 15 (a) 
 supestes, 89 
 suppingo, 16 
 supra, 15 {a) 
 sups-, 91 (a) 
 suptilis, 64 
 surripio, 10(3) 
 surrupui, 10 (3) 
 Buscipio, 91 (a) 
 suscito, 10 (1) rt 
 sustineo, 1, 10 (1) b 
 Sustus. 58 
 sutor, 27 
 
 taberna, 10 (3) 
 i taceo, 10(l)c 
 
 tf» 
 
 M 
 
 I 
 
 f) 
 
 - tam, 46 (2), 80 
 
 *' f * y tamquam, 80 
 
 tan, 46 (2) 
 
 tango, 10 (1) c 
 
 tanquam, 80 
 
 ted, 38 
 
 tego, 1, 33 (i., iv.) 
 
 tegula, 33 (iv.) 
 
 tela, 91 {b) 
 
 temperi, 14 
 ' . templum, 46 (1) 
 
 temporis, 14 
 
 tempus, 14 
 
 teneo, 1, 10 (1) b 
 
 tesa, 87 
 
 testis, 89 
 
 tetigi, 10 (1) c 
 
 tetuli, 15 (6) 
 
 texo, 91 (b) 
 y thensaurus, 87 
 
 thesaurus, 6 (iv.) 4, 
 87 
 
 Thorius, 35 
 
 tibi, 32 
 
 tibicinium, 10 (1) c 
 
 toga, 33 (i.) 
 tondeo, 85 (1) 
 tonsor, 43 
 m tonstrix, 85 (1) 
 
 • Torius, 35 
 
 tormentum, 90 
 torqueo, 90 
 torreo, 74, 89 
 tostus, 89 
 tot, 31 (1) 
 totidem, 31 (1) 
 tracto, 13 (1) 
 tradere, 11 
 tradidi, 10 (1) b 
 traduco, 92 
 trameo, 92 
 trano, 92 
 transduco, 92 
 transfundo, 13 (3) 
 transigo, 10 (1) c 
 cransmeo, 92 
 \ j^l i» transno, 92 
 
 ' "^ ■ ' tredecim, 86 
 
 triumpus, 35 
 
 r," 
 
 V^ 
 
 INDEX 
 
 Troiad, 38 (2) a 
 trucis, 29 
 trux, 29 
 tuber, 40 
 turn, 80 
 tunc, 80 
 tunde, 31 (1) 
 tundo, 30 
 turturis, 29 (a) 
 turtur, 29 (a) 
 tutudi, 30 
 
 U 
 
 ulciscor, 90 
 ultus, 90 
 umbra, 46 (1) 
 umerus, 44 
 unctio, 82 
 unctus, 90 
 unguentuin, 82 
 unguo, 90 
 uni, 27 
 unius, 27 
 unum, 21, 46 (2) 
 unus, 21 
 urbi, 64 
 urbis, 64 
 urbs, 64 
 urps, 64 
 usus, 21 
 
 vafer, 40 
 valde, 15 {a) 
 Valerii, 41, 43 
 Valesii, 41, 43 
 validus, 15 («) 
 vapor, 43 
 vapos, 43 
 varietas, 12 
 varies, 12 
 varius, 12 
 vehemens, 35 
 velira, 17, 33 
 velle, 17, 65, 74 
 vellem, 17 
 
 107 
 
 venalis, 45 (1) 
 vene, 36 
 Veneres, 29 {b) 
 Veneris, 29 (6) 
 veni, 36 
 venos, 13 (2) 
 venustus, 13 (2) 
 verber, 11 
 verberis, 11 
 verres, 74 
 Vesuntionem, 36 
 via, 47 
 
 vicis, 10 (1) a 
 viden, 31 (3) 
 video, 10(1) a, 26 
 vija, 47 
 vilicus, 59 (1) 
 villa, f)9 (1) 
 villum, 72 
 vincio, 90 
 vinclura, 45 (1) 
 vinctus, 90 
 vinum, 72 
 
 vir, 29 
 
 viri, 29 
 
 viro, 46 (2, 4) 
 
 viro(m), 29, 46 (2) 
 
 virum, 29 (d), 46 (2) 
 
 visus, 26 
 
 vivont, 29 
 
 vivunt, 29 
 
 vixi, 36 
 
 voco, 14 
 
 volam, 17 
 
 volebam, 17 
 
 volneris, 11 
 
 volnus, 11 
 
 volo, 17, 33 
 
 volui, 17 
 
 voluptatem, 31 (3) 
 
 vomer, 17 
 
 vulnus, 17 
 
 Zanuariu^, i^S 
 zephyrus, 6 (iv. ), 3 
 zophyrus 6 (iv.), 3 
 Zulia, 53 
 
\ 
 
 'M 
 
 « 
 
 
 PRINTED AT THE EDINBURGH PRESS 
 9 AND 11 YOUNG STREET 
 
 S.JJ. 
 
 1 ^ 
 
 I 
 
 -.■*!« 
 
 jg -^ -S-^A Vfe, -t 
 
877.15 
 
 N555 
 
 COLUMBIA UNIVERSITY LIBRARIES 
 
 This book l3 due on the date hidlcated below, or at the 
 expiration of a definite period after the date of borrowing, as 
 provided by the library rules or by special arrangement with 
 the Librarian in charge. 
 
 OATC BORROWED 
 
 DATE DUE 
 
 DATE BORROWED 
 
 DATE DUE 
 
 
 
 
 
 
 ^m T^ 
 
 
 
 
 Kun< f Fft 1 
 
 5t9ai 
 
 
 
 Lr\ni» % *-*- < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 C28 (449)M50 
 
 
 
 1 
 
 1 
 1 
 
 f 
 
 il 
 
 
 
 \ 
 
 h \ 
 
 f 
 
 %ll.l^ 
 
 APR 
 
 •i r., \3A^ 
 
 >♦ 
 
 NXT3 
 
 
 t 
 
 (Si 
 
 m 
 
 »C2s 
 
 m \ 
 
 2^1 
 
 R-ioe 
 
 % 
 
 
 c 
 z 
 5 
 
 < a 
 
 -J 
 
 " M 9n iq4q 
 

 .1^ 
 
 
 
 i-' 
 
 
 1-^ 
 
 ^.i 
 
 '3 
 
 ;:S-5> 
 
 »•• 
 
 ,-3; 
 
 ,.-,, 
 
 ■;-»■ 
 
 
 ■K,-s 
 
 -'A-!