An Analytical Essay on the Greek Alphabet  

From The Art and Popular Culture Encyclopedia

Jump to: navigation, search

"In 1791 Richard Payne Knight published An Analytical Essay on the Greek Alphabet, in which he argued that Michel Fourmont had forged some inscriptions in his collection."--Sholem Stein

Related e



An Analytical Essay on the Greek Alphabet (1791) by Richard Payne Knight.


Full text




HE Subje&, which I here propoſe to examine, will of courſe ap- pear minute and frivolous to thoſe, who are only acquainted with it from the keen ridicule, with which it has been treated by ſome popular and elegant writers of the laſt and preſent centuries (1). I would, however, entreat all perſons of this deſcription, who honour the preſent attempt with their attention, to conſider, that even the beſt and keeneſt ridicule is no teſt, either of the truth or the dignity of the ſubject, upon which it is em- ployed, but has often been moſt happily exerciſed upon the beſt- founded opinions and moſt important and elevated objects (2), At all events, I hope that they will not condemn the deſign before they know the conſe- quences of its completion ; and if they then find that, by facilitating the acquiſition of Gretian Learning, it can bring the higheſt efforts of hu- man taſte and genius, into a ſtronger or clearer light, they will conſider it as adding to the intellectual pleaſures of man, which are certainly the moſt valuable belonging to his nature, becauſe they can be at all times en-

Joyed without injury to dann, $200 or fortune.

1013 + 

(1) See Moliere's Bourgeois 8 42 Pope's Doscial. (2) See Gulliver's Travels; and Tale of a Tub. r

5 B . . TI cannot


I cannot indeed but think, that the judgement of the Publick, upon the reſpective merits of the different elaſſes of Criticks, is peculiarly partial * unjuſt.

Thoſe among them who aſſume the office of pointing out the 8 and detecting the faults, of literary compoſition, are placed with the orator and hiſtorian in the higheſt ranks ; whilſt thoſe, who undertake the more laborious taſk of waſhing away the ruſt and canker of time, and bringing back thoſe forms and colours, which are the ſubje& of criticiſm, to their original purity and brightneſs, are degraded, with the Index-maker and

Antiquary, among the pioneers of literature, whoſe buſineſs it is to clear

the way for thoſe who are capable of more ſplendid and honourable enter- prizes.

But nevertheleſs, if we examine the effects produced by theſe two claſſes of Criticks, we ſhall find that the firſt have been of no uſe whatever, and that the laſt have rendered the moſt important ſervices to mankind. All perſons of taſte and underſtanding know, from their own feelings, when

to approve and diſapprove, and therefore ſtand in no need of inſtructions

from the Critick ; and as for thoſe who are deſtitute of ſuch faculties, they can never be taught to uſe them; for no one can be taught to exert facul- ties which he does not poſſeſs. Every dunce may. indeed, be taught to repeat the j jargon of criticiſm, which of all jargons is the worſt, as it joins the tedious formality of methodical reaſoning to the trite frivolity of com- mon-place obſervation. But, whatever may be the taſte and diſcernment of a reader, or the genius and ability of a writer, neither the one nor the other can appear while the text remains deformed by the corruptions of blunder- ing tranſcribers, and obſcured by the gloſſes of ignorant grammarians. It

is then that the aid of the verbal Critick is required; and though his mi-

nute labour, in diſſecting ſyllables and analyſing letters, may appear con- temptible in its operation, it will be found important in its effect.

The office, indeed, of analyſing letters has been thought the loweſt of all literary occupations; but nevertheleſs as ſound, though only the ve-

hicle of ſenſe, is that which principally diſtinguiſhes the moſt brilliant

poetry from the flatteſt proſe; and as, in the dead languages, all ſound is to be known only from the powers originally given to the characters re- preſenting the elements of it; to analyſe theſe characters, and ſhow what



their Powers really were, is the only way to acquire a knowledge of thoſe ſounds in which the antient poets conveyed their ſenſe. A ſucceſsful en- deavour to obtain this end will not, I flatter myſelf, be * either teifling or abſurd in this age of taſte and in

N articulate ſound is properly that which begins from, or ends in,

a ſuppreſſion or obſtruction of expiration, by the compreſſion of ſome of the organs of the mouth. _ "Theſe organs are the lips, the teeth, the tongue, and the palate; to' which ſome add the throat, but improperly, for guttural ſounds are not of themſelves articulate : the combinations of them known to the Greeks were only three; I. the lips with each other; II. the tongue with the pa-

late; III. the tongue with the teeth: to which the Latins added a fourth, of the under · lip with the teeth: but this the Greeks never employed, and therefore could not pronounce the Roman F (1), though we perpetually pronounce it in our corrupt manner of reading their language. To repreſent theſe three modes of articulation, I am inclined to believe, the firſt viſible ſigus for ſounds were invented; for, though articulation be only the form, and tone the ſubſtance, of ſpeech, yet as the form is finite and ſimple, and the ſubſtance infinitely variable, it is natural to ſuppoſe that the firſt ſigns were invented to repreſent form rather than ſubſtance, It is alſo this form or articulation which diſtinguiſhes human ſpeech from the cries of animals, which are all tones, or vowel ſounds, variouſly aſpi-

0 See Quintil I. xii. c. 10. E 2 B 2 rated, a


rated, but neither begun, ended, or divided, by the compreſſion of the organs of the mouth,

The firſt ſigns or notes of articulation were, therefore, the G (as it was antiently pronounced, and as we till pronounce it when followed by an A, O, or U), the P and the T (1).

Each of theſe was pronounced two ways, with a greater or leſs degree of force in the compreſſion of the organs; whence were formed three more letters, B, K, and D, which I rank next in ſucceſſion, though there is reaſon to believe that nzither of them (or, at moſt, only the laſt) was in- vented until ſeveral intermediate improvements had taken place in the art of expreſſing ſounds by ſigns. The want of authentic monurrents, how- ever, prevents us from tracing the- progreſs of theſe improvements, the earlieſt inſcriptions extant having been made when the Alphabet was even more perfect than it is at preſent. It ſhould ſeem, indeed, both from the order of the Alphabet, and our manner of pronouncing theſe letters, that the B, G, and D, ought to rank together in the firſt claſs; and the P, K, aud T, in the ſecond ; which would certainly agree better with the ana- logy of ſound; but, nevertheleſs, it is contradicted by the authentic teſti- mony of antient monuments, always to be preferred to any concluſions that can be drawn from mere analogy.

In a very antient Greek Inſcription found. in Magna Græcia, and now preſerved i in the muſeum of Monſignor Borgia, at Veletri, the G is ex- preſſed by a ſingle perpendicular line, thus I (2), which ſeems to be its moſt antient form; for, upon ſome of the earlieſt coins extant, it is ex- preſſed by the ſame line a little curved, thus ( (3) ; whence came the Ro-

man C, which is uſed for the G in the Duillian inſcription, engraved in the year of Rome 493. The G was not employed. as a diſtinct letter until introduced by Spurius Cervilius Ruga, twenty-ſeven years afterwards (4). Antiquaries have obſerved that, in Manuſcripts, the round forms moſtly

(1) I employ the Latin letters becauſe much nearer to the primitive Greek than the Greek ones now in uſe,

(2) Plate 1, Fig. 1, from a copy of it given me by Mr. Aſtle.

(3) See thoſe of Gela in Numm. Sic, vet. Pl. XXXI.

(4) Plutarch. Qy. Rom. Taylor's Civil Law, p. 5571 alſo, in Marm. Sandvicenſ.



predominate in the letters, and in inſcriptions the ſquare, becauſe the for- mer are more eaſily written, and the latter more eaſily carved (1). Hence this curved Line, which repreſented the G, was made with an angular in- ſtead of a circular curve, thus (, or thus F.

The moſt antient K is a combination of one of theſe forms with the ans tient upright line, thus J. or thus > ; fo that this letter is, in fact, a junction of two Gammas, in order to expreſs a ſtronger and more empha- tical enunciation by the ſame organs. This will appear evident by exa- mining the manner in which it is repeatedly written in the Etruſcan In» ſcription, called the Eugubian Table, publiſhed by Gori; and alſo upon ſome very antient medals of Leſbos and Syracuſe, in both of which it. is plainly repreſented by two diſtin characters (2). This Etruſcan Inſerip- tion Gori endeavours to prove, from a paſſage of Dionyſius of Halicarnaſ- ſus, to have been written two generations before the Trojan war; but, though I do not think his argument quite ſatisfactory as to this point, it is of very remote antiquity, for the Alphabet is the moſt imperfect, and therefore, probably, the Med of any hitherto diſcovered.

Upon ſome very —— coins of Croto, Corinth, and Syracuſe, we find the Kappa expreſſed by a circular ſupported by a perpendicular line, thus O (3), from which comes the Roman Q. This is, however, equally a combination of the antient Gammas, the two curved lines being Nes and divided by a perpendicular one, thus 0,

After the invention of the Kappa, the ſimple Gamma ſeems to have fallen into diſuſe in ſome dialects; for it is not to be found in any Etruſcan inſcription ; and the Etruſcan, as well as the Latin, is evidently a corrupt dialect of the Greek; a dialect by much the harſheſt of any, and therefore probably employing only the barſheſt and moſt emphatical palatial conſo- nant, which is the Kappa,

Both theſe letters retain their powers, with, I believe, little or no vari- ation, in moſt of the modern a "m_— that the Engliſh, F n

(1) See Aſtle's Hiſtory of Writing. (2) See Plate I. Fig. 4 and g, from coins in the cabinet of the 38 8 (3) See Comb, Pl. AX. XXI.; and Torremozzi Sic, Pl. LXXVII. Similar medals are in the cabinet of the Author; who has quoted none that he has not ſeen, having too often proved the inaccuracy of books in theſe minute but important circumſtances. 9 55 and


and Italians, have added a corrupt and barbarous dental ſound to the G, when followed by either of the ſlender vowels. The K is not employed by the Italians, Spaniards, or French, in their own tongyes.; and in read- ing the Greek they pronounce it in the ſame manner as they do the Latin C, that is, like a barbarous ſemi-vowel, forced out .between the MET and the teeth with : harſh hiſſing ſound.

Ihe moſt antient form of the P ſeems to be that of the Etruſcans, which conſiſts of a perpendicular line with another drawn obliquely from it, thus 4. It exiſts in the ſame form, except that the oblique line is curved thus (J, to diſtinguiſh it from the antient Lambda, upon the vaſe repre- ſenting the hunt of the Caledonian boar, in the Britiſh Muſeum, which is evidently Greek, and appears, both from the ſtyle of the workmanſhip and form of the letters, to be one of the moſt antient monuments extant of the art of that people. This curvature, being gradually increaſed, formed the Latin P, which was previouſly employed by the Greeks in the ſame form, as appears from the very antient Veletrian Inſcription before cited, In the ſame inſcription, however, it appears in the form which they more commonly employed in early Times,; which is indeed nearly the ſame, only that the curved line is made ſquare inſtead of round (17), for the rea- ſon beforementioned. The power of this letter ſeems not to have varied at all, for it is preciſely the ſame in all the languages of modern Europe, and, as far as we can judge from analogy and etymology, the lame as It was in Greece in the days of Homer,

The B ſeems to have been originally an aſpirated P; for, in the Eugu- bian Inſcription, it has that power; and the Macedonians employed it

where the Greeks employed the @ and N, writing BEPENIKH for @EPE- NIMH, and BYPDOL for IIYPDOE; whence it appears that our Northern words BURGH and BEAR come from the ſame ſource as the correſponding

ones in the Greek. The Etruſcans repreſented. it in two forms, thus H, and thus 8; the firft of which occurs only once, and that in the Eugu- bian Inſcription ; but the other is common, It is with the firſt that both the Greek and Latin forms of this letter agree; but its power ſeems to have been that of the Phoenician Beth, at leaſt if they pronounced it as we do now, which the Greeks ſeem evidently to have done in ſome ju- ſtances ; for the verbs BOMBEQ, BAMBAINQ, &c. would not have au- ſwered


ſwered the purpoſe for which Homer employs them, in making the found correſpond to the ſenſe, if the B were pronounced in any other manner. In other inſtances, however, or, at leaſt, in other times, they employed it as a palatial aſpirate; for we find the Latin V (which we know had the power of our W) ſometimes expreſſed in the Greek by the B, and ſome”

times by the OT diphthong (1) ; whence it clearly appears. that there was then an affinity between them, though they now differ ſo widely, The ZEolians and Dorians, in particular, employed it occaſionally. as a pure or ſimple aſpirate, like the Digamma, or Roman H, writing BPOAOZ for

POAOE, BABEAIOE for AEAIOE, BEAOE for EAOE, &c. (2). In the 

ſame manner it was introduced into the words TAMBPOL for TAMEPOYL, and MELHMBPIA for MEZHMEPIA3); but with what degree or form of aſpiration it was pronounced it is impoſſible for us now to tell ; for though, like the OY diphthong, it had a reſemblance to the Latin V, we cannot ſay how near that reſemblance was. In all modern languages it retains its antient power of a labial conſonant, except in the Spaniſh, and ſome dia- les of the modern Greek, in which it has acquired that corrupt and bar- barous ſound' given by the other nations of Europe to the Latin V, a ſound which it ſeems to have derived from the Byzantine Greeks, as it is en- forced by the edi& ifſued by Stephen Gardener, Biſhop of Wincheſter, .for the ſupport of their pronunciation in the univerſity of Cambridge, of which he was Chancellor. The Romans: ſeem to have been very licentious and irregular in the uſe of this letter; for on the Duilian column, before al-

luded to, the name, which in later times was written DuiL1us, is writ- 7 ten BI. los; whence, as Gori obſerves, BELLUM and BELLONA appear to be the ſame words with DUELLUuM and DUELLONA (4) ; and we find ac- cordingly, in the Senatus conſultum Marcianum, inſcribed about ſeventy» five years after, the name of the goddeſs BELLONA written DVELoNA. In the inſcription in honour of L. Scipio Barbatus, which 1s of the year after the Duilian, the B is alſo ropreſented by the 1 and V in pvonoRo, the

(1) As in the names Vanzo and SevERYs, e written by Greek authors BAPPAN and ZEBHPOZ, and ſometimes OTA PN and ZEOTHPOE,

(2) Priſcian, lib. I.

(3) Lennep, Analog. Græc. p >. 286,

(4) Mu. Etruſe. Claſſ. V.

  • . Car os


antient form of the word nowoRuUM, the final M having been uſually omitted, and the U repreſented by the O in the old Latin.

The moſt antient figure of the T, found in the Etruſcan inſcriptions, differs little from that now in uſe. Its power has alſo probably continued the ſame, except in the inſtance of the hiſſing ſound, which moſt modern nations have given it, when followed by an 1 in the ſame ſyllable, This is undoubtedly a corruption, the Greeks having no letter to expreſs thus kind of ſound but the Sigma.

The D, the other dental conſonant, does not appear to have been known to the Etruſcans, having been probably borrowed from the Phoe- nicians after the Pelaſgian alphabet had been carried into Italy. Its figure, indeed (which is always triangular, though often rounded at one angle), occurs frequently on the Etruſcan monuments ; but it always ſtands for the R. We find it, however, with the power of the D, or perhaps the Ax or z, upon the Zanklean medals, which contain ſome of the moſt an- tient ſpecimeus of Greek writing now extant (1).

Theſe ſix letters are called mutes, becauſe, if employed according to their original intention, they expreſs no ſound of themſelves, but only mark the beginnings, endings, and diviſions of ſound, by which it is arti-

culated, or ſeparated into detached portions, called in writing ſyllabtes, 

Theſe portions are, however, often divided by other means, which 1 ſhall now proceed to examine; but, in that cafe, it will appear that they are not, ſtrictly ſpeaking, articulate Rants or eſſentially different from the cries of brute animals.

The firſt of theſe is a partial inſtead of a total ſuppreſſion of the breath, by an approximation inſtead of a conjunction of the organs of the mouth, repreſented by the letters called aſpirates; which, like the mute conſo- nants, are to be divided into three clafles, correſponding to the three dif-

ferent combinations of the organs of ſpeech. 

But, as each of theſe marks ſignifies a particular mode of conſtrained ex- piration, by the approximation of ſome particular organs to each other, the moſt natural and eaſy way of expreſſing them would be to invent ſome

(1) See Torremuzzi Sic. Pl. XLV. Similar medals are in the Author's cabinet, and in moſt others, they being common.

7 7 mark

\ \ 5 4 » * < * =

- * 4 — — — — 92


J : OS 0 12 A y 0&4 & :

  • * om —„œô. u 2 „ „„ „% A ore ans 2 P g p

6 D —— — — 0 + wo W ů o˙ 2

; 7 -   4 — — ä 4 AD 4h 

- = o . * * . * N 3 4 : 4 + 0 © * U bs -

= of ; - ” ſn — = TY © 4 . ec e . * FF by

2 , = . 2 k

  • * = N N ö \

-. f - - ö ö a

  • — —— ————_——— : : .

- = 2 — - * —— — — OUS I ooo —— «„ 2 K <5 — "_

— — -— —- 0 —— *
—_ — — —— . 

7 1 ,

  • * F


# % - ” i 7 1

0 5 = * 1 8

= — — — N — — — — — — — 2 — © — - - — — ———— — — — — — —— ——— —⅛— n. — — — „ . — en 2 --- * —

= — —2— > — 

0 *


P =

ts : - - avant. ' 7 — —— — —-—-—¼ 32 K

- &

— 1 N * _ as ud 4 P_ — ————C ww ——w—U—U—[U—ä—— ———— — out. — Hos at — —— P— an n _— 27_————— « , F £ - . ' . 4 2 f 2 „ K % 00 2 — — ——— UU —— — 0 7 — - — r — ————— 8 N 0

; ” —n—— OS 

. - - ,

4 - - - a , 2 - . n - - -

0 1 — —ũU— — 4 — on — — n ö — — —— — 2 — —— 2 — —ͤ— 5151 — © Ae — - -0 £ — „„

r —

- —_— . ——— 0 2 = — m . - - . 2 = - - % — LR - ww =. os -_ - — 4 — „„ ws — - * - - „ „ 2 „ 4 4 — — — *

  • 1 *

A =-

. 2.0 / ig F £2 «  ; . . — — Jr r—_— — 22 x A 2 — — —— — —— — —— 2 — — — vos — ——_— — —— . wo. — -- . — -# - - 1 n 1 S 2 2 —_— * - _— - PR *

  • 4 ;

- M

  • r

2 4 1 — _ * „— 2 244 - =

* - o

1 - — — — — — — — — — - * 1 1 F *

  1. *

. - ' - -

-. . * : 1 — — 1 4 — — 6 „„ ͥ ꝗ ꝗ „ — eons „ — —ͤ—ñ—— ́ F —+ — — — — - — — 2 „„ „ „ *** * 1 Is * 0 —ͤ—ͤ—ä— 4 ws * "_ OO . Ns cn — *


f i * * 4 a

' .

 "4 = 

At " — _ —

  • ” *

4 „ -

  • 2 4

„ - . 4 b . — = oo ks 44 4 * 2

1 a - „ _ * ä 1 _— — —— — ñ ̃mſ1— . — - » —— — ”One öð»òö O R we a 6 + we oo * — —2—2—¶ᷓ0 9„«„ — öꝗ—ũ—ñ ” cas” mw » a - — —— « PIETY Wa 0 ä — Py" _ — —-—̃— - * 2357 8 a = . ”- - . # * : [ - a _ RO *

  • _ *

" - 3 4

  • 9 N = Y

2 - . ”

- 2 1 * — .

P p . 7 

7 - 4 . 8 r P 7 - ( - 4 — - E y 2

  • .
  • - P 0 PR

— th. Mt — „ — * 4 — — —— — „% „% ĩ ͤ K70%69„%c„„ — — 7 - 7 7 , 1 þ ; ; . - „ ew cue id» < — Ws „ „ — —2 a —ͤ—ͤ EU, Cs cout tt tes — — — 2 GO 4 2.286 * S8 „„ a 4 * I. Roos x gg. frac Ee - „ 23 — 2 ——⏑+–˖ 9 —_——— 4 \ > 7 4 ; „

  • 1 4 -

. — 3 * 1 eg;

  • — ꝗ— m — * © da 'wY

8 -v * - - N N P


— —


N Mb

+ ?

3 —


ve —


adv te) Che 57

G.< «„ „




TREE We — 2 1 A VV 7. * 932 by Þ - ov ON — GAU wr Bon

fart 71ofes of arkeulaher. N Rear C feb akin

— Fl — . — > —— ͤ—— — — —— — —

— — —

  • — 22 2 —2 —— - - — — — — —

DA AI T's — — _— — — — — — W 7 —— == — — — — — —

— 4. — —— — — —— — — — — — — ——— — —U— — — —ͤ—— —— —_ — — » —— LE. — —

0 .

F © ga: 3 Gao may wag formed /


— e e ASS — —ñ —-—-k— — — —— —̃ — — N. 7 — — —— $— = 1 — —— — — — — — — — — — — — — — — K ones te

3 — P. 1.9 1 ,

— — ꝗ— 2 *





— *

- 0 „ 4 wal - — —— — 2 — 2 —— 22 — . — jo _ _ *

— ä—ꝶä—Hͤ — 2 — — ——ä— Cur oo * —  



ON THE GREEK ALPHABET. 9g mark of general conſtrained expiration, which, being affixed to each of

the ſigns before invented, might diſtinguiſh each different mode of con-

ſtrained expiration according to the different combinations of the organs by which they are produced. Hence come the ſimple aſpirate, figured by the Phcenicians and Etruſcans thus U., by the Latins thus H, and by the Greeks thus H, and thus F; which, being prefixed to a vowel, ſignifies that the tone, which it expreſſes, ſhould be uttered with a forced and con- denſed expiration; and, when affixed to a conſonant, that the breath, which forms that tone, ſhould not be totally ſuppreſſed and interrupted by it, but only confined and conſtrained by the approximation only of thoſe

organs, the entire junction of which is ſignified by the conſonant alone,

The ſecond Greek charaQter for this ſimple aſpirate does not ſeem to have been in uſe till the other was appropriated to expreſs another letter. An antient ſcholiaſt, cited by M. de Villoiſon (1), ſays, that, when the H became a vowel, it was divided into two letters, the firſt of which, F, was employed to ſignify the aſpirate, and the ſecond, I, the flender, or ſimple vowel ſound. Quintilian and other old grammarians' ſeem to have held the ſame opinion (2); ſo that there can be no doubt but that theſe marks were ſo employed in the manuſcripts of their times. There is, however, no inſtance of the I in any antient monument now extant, or in any manuſcript anterior to the ninth century, though the F occurs upon the medals of Tarentum, Heraclea, and Leſbos, and alſo on the He- raclean'tables, and an earthen vaſe publiſhed with them by Mazochi; who has conjectured, with much ingenuity and probability, that theſe two notes were firſt employed in oppoſition to each other, to ſignify the thick and ſlender enunciation of tone, by Ariſtophanes of Byzantium, the inventor

of the accentual marks (3). The preſent notes (c) and () are corruptions

of them, which were gradually introduced to facilitate writing (4). Dr, Taylor ſuppoſed that the * was the Ionian W the F the Dorian,

(1) Proleg. in Homer, p. * whers the marks, through. an error of the n or er,

are tranſpoſed. (a) Lib. I. c. 4. & Gramm, vet. Putch, Col. 1829, & /q.

(3) Comm. in Tab. * *  » eh (4) Ibid, *

. e aud


and the F the olean (1); but we find the F in its Pelaſgian Form, C, with the F on the Heraclean tables; and the Leſbians, whoſe coins have the latter aſpirate, which he calls Dorian, were /Eolians,

Diſtin& marks or characters were invented for each of the aſpirated con- ſonants at a very carly period; fo that, I believe, there is not more than one genuine example extant in which they are ſeparated in the primitive mode. This is a votive inſcription preſerved at Venice, in which we find KH for x, and Un for o, as in the Latin (2), which was derived from the Æolian or Arcadian alphabet, before the aſpirated conſonants had found a place in it, In the oldeſt Etruſcan Inſcriptions, however, as well as the Sigean, ſuppoſed to be the oldeſt Greek extant except coins, we find them, both pa- latial, dental, and labial, expreſſed by characters not only diſtinct, but which have no apparent reſemblance of form to the letters from which they are derived.

The palatial aſpirate, which conſiſts of either the Gamma or the Kappa aſpirated, was made by the Etruſcans, I believe, invariably, and by the Greeks ſometimes, like a divided V, thus W(z). Its uſual form, hows» ever, was compoſed of two tranſverſe lines thus X; which, on the very antient medals of Naxus in Sicily (4), is employed, as in the Latin, to ſignify the & or abbreviated mark for the PE and KZ, unleſs indeed, as I am inclined to think, the name of that city was really NAXLZOE contrafted to NAXOZ, as ALANKAE to AANKAE, by an eliſion of the T, much af-

feRted by the Greeks in the refinement of their language, when the ſound 

of that letter was deemed harſh and barbarous. The power of the Greek X ſeems to have been nearly the ſame as that which the Spaniards now give to the Roman X, the Tuſcans to the C, and the Scotch to the GH. We are apt to pronounce it as if it were a plain K without any aſpiration ; and the French have given it the barbarous ſound of their own CH, a ſound which to a Greek would have appeared ſcarcely human, It was pro-

(1) Ad Marm. Sandvicenſe, p. 4

(2) See Pl. I, Fig. 2. I have not ſeen the originals, nor any fac: ſimile either of this as the Veletrian Inſcription ; but as both have been generally acknowledged to be authentic, and contain no internal evidence to the contrary, I have ventured to quote them,

(3) See Pierres gravees du Duc d' Orleans, Tab. II. Pl. II.

(4) See Torremuzzi, Pl. III. Fig. a, from a medal now in the cabinet of the Author. 1 bably7


bably pronounced more or leſs gutturally i in different dialects, or 3 as it was compoſed of the F or K, the latter of which letters was ſome- times employed alone as a palatial, and the former as a guttural, aſpirate, The F in particular was prefixed to words in ſome dialects, and omitted in others, as the aſpirates frequently were; whence Homer writes FTAOYIIOL and AOYAOE, DAIA and AlA, &c. as the metre requires. Hence too we may perceive that the Latin cum and the Greek EYN are the ſame word, the original form of which was FZ TN, now written Fu, from which the one nation dropt the ZE, and the other the P. This is the rea · ſon alſo that in the Latin the S is frequently prefixed to another conſonant without rendering the preceeding vowel long, From this uſe of the Gamma probably came the Digamma ; which, from its form as well as name, ſeems to have been compoſed of two Gammas placed one upon the other thus f, or thus L; the former of which figures was employed by the olian "6 Ionian Greeks ;. and the latter by the Etruſcans, Campanians, and other Pelaſgic clans of Italy. The Latins retained the Greek figure in their Alphabet, derived from the Arcadian, which was alſo the Æolian; but they corrupted the ſound of it in a manner that is difficult. to be accounted for. The Digamma was cer- tainly pronounced rather as a ſimple aſpirate than as an aſpirated. conſo- nant, and differed from the common note of aſpiration in the impulſe, which cauſed the forced expiration, being given from the throat rather than from the tongue and palate: but the Roman F was pronounced by a

forced expiration from the under-lip through the intervals of the upper

teeth, ſo as not to reſemble any voice, whether of man or animal, according to the obſervation of Quintilian (1). It is generally ſuppoſed among the Learned at preſent, that the Digamma was pronounced like our W, for it correſponded to the Latin V, the ſound of which was certainly the ſame. The etymology of many Latin words proves this; vis, vicus, VINUM, &c. being evidently from Fiz, FOIKOE, FOINON, &c. the two laſt of which were probably once written FiKOE and FINON, whence our words WICK and WINE ; for, upon the very antient medals of Oaxus in Crete, we find the O omitted, and the name of the city written FAZOZ (2). In the Veletrian Inſcription it is however inſerted in the word FOIKOZ, The

(1) Lib. XII. c. 10, © (2) See Dutens, p. 165.

W, as




W, as pronounced by us, is a palatial aſpirate of the flendereſt kind, hav-

ing more of tone than articulatjon, and being rather a vowel than a conſo- nant, for it is uttered with little or no conſtraint of expiration. The Welſh commonly employ it to expreſs tone only, with conſonants, as we do in ſome inſtances, though always accompanied by another vowel, as in TWINE, TWIST, DWELL, &c.; in all which the W is as much a vowel expreſſing tone as the I or E. The difference, however, between a pala- tial and guttural aſpirate is very ſmall ; for, if the tongue and the palate are a little more than ordinarily compreſſed, while the breath is forced be- tween them, the compreſſure naturally extends to the throat, and the ſound becomes guttural. Local or temporary habit is always ſufficient to cauſe this; wherefore the ſame letter, which in one age or province was employed as a palatial, might in another have been employed as a guttural, aſpirate, The Zolic dialect, we know, had more guttural ſounds than any other, and more particularly employed the Digamma, which 1s thence

\ ealled Zolic by the later grammarians. We may, therefore, fairly conclude that it repreſented this ſound, to which, perhaps, there is nothing nearer

in modern language than our wn, as pronounced in the word wHIRL ; or that of the Tuſcan ou, as pronounced by the natives of Florence and Piſa in the word unn RA. The Pelaſgian vau, from which is derived the Roman V, had certainly the ſame power, and was often confounded with . it; and we know that this letter was an aſpirated Y, en which the vowel 7 was diſtinguiſhed by the epithet þ:a0y.

Both the F and the H or F ſeem to have been dropt from the Greek Alphabet nearly at the ſame time, probably about the period of the Perſian

war. The firſt figure of the latter was, however, retained, to repreſent

the double or long E, and the former ſeems to have continued in uſe in

particular places, where a fondneſs for the antient dialects prevailed, even

to the final ſubverſion of the Greek republicks by the Roman arms. Strabo ſays, that the people of Elis and Arcadia preſerved the Æolic dialect pure

When it was mixed or loſt in every other part of the Peloponneſus (1), and 

of courſe in every other part of the world. In collections of antient coins we find a great many inſcribed FA and FAAEIQN (2); ſome of them

0) Lib. VIII, (a) See Comb, Pl. XXVII. Fig. 21, 24, 24+



ſtruck at the earlieſt period of the art; and others apparently under the Achizan league, as they are of the lateſt ſtyle of workmanſhip, and have the uſual device of that federative republick imprinted upon them (1).

FAAEIOI we know muſt be the olian manner of pronouncing HAEIOT, the people of Elis, to whom, I bave no doubt, that theſe coins belong, and not to the Faliſci, a people of Italy, to whom writers upon medals have 1gnorantly aſcribed them, without "conſidering that neither the letters nor inflexion are ſuch as couls have been employed by the antient POIs of Latium or Etruria. is

The labial aſpirate © was uſually repreſented · in the Etruſcan alphabet by two circles one above the other like the Arabic figure of eight (2). In the Sigean Inſcription it is of the form now employed (3), which has ſcarcely ever been varied, except in making the interſeQed circle ſquare for the convenience of engraving. It was pronounced antiently by à con- ſtrained expiration between the lips, which approached towards each other; but all the modern nations of Europe pronounce it hke' the Roman F, though that was a letter which the Greeks were abſolutely incapable of uttering, there being no ſound in their language which at all reſembled it; Hence they were abſurdly and illiberally ridiculed by Cicero for bringing an accuſation againſt Fundanius when they could not pronounce his name (4). For the credit of modern manners, I believe there is no court of judicature now exiſting that would liſten to ſuch a defence, if an advo- cate ſhould be ſo forgetful of decency as to attempt to employ it,

The ©, or dental aſpirate, was repreſented both in the Etruſcan and Ionian alphabets by a circle interſected by one or two tranſverſe lines, thus O, or thus O: or having a point in the middle, thus O. The antient manner of pronouncing, it was indiſputably that which is ſtill obſerved by

the modern Greeks, the Copts, and the Engliſh; that i is, by a conſtrained

expiration between the tongue and the upper teeth. All the other Euro- pean nations pronounce it as a mute conſonant, and throw the aſpiration

upon the next Minesse kn vowel, This is a- ſort of e egd defect; 5

K 3.* 4119

(5) See Geſner. pl. XIV. Fig. 73 alſo, KY p- 5+ U an K 108 1 Di (2) See Eugubian Table before cited, 1 : (3) See Pl. II.

00 Quintil, lib, XII. c. 10 bug. antiently



antiently the Northern nations could not pronounce any of the aſpirated conſonants; whence, among the barbariſms uttered by the Scythian, in the Theſmophoriazuſz of Ariſtophanes, we invariably find dhe Kk for the Xx, the I for the o, and the T for the 9

F The 2, called San and Sigma, which is found under different forms in all alphabets, and which grammarians claſs ſeparately by itſelf, as being

neither mute, aſpirate, or liquid, is in fact a dental aſpirate, differing from

the & only in being pronounced with the tongue applied to the root in-

ſtead of the point of the teeth, ſo as to produce a hiſſiug, and what ap- peared to the refined ears of ths Greeks, a barbarous ſound. This hiſſing

pronunciation of the dental aſpirate ſeems to have been the only one known

1 to the Lacedæmonians; for, when brought upon the ſtage by Ariſto-

phanes, they uniformly uſe the L for the G. It appears, however, to have 

been only a local and vicious habit of pronouncing ; for, had it been an

eſtabliſhed characteriſtick of their dialect, we ſhould have found the ſame

ſpelling in the treaties of alliance entered into by them with other Dorian

States, which are always in the Doric diale&, but without this peculiarity.

In other inſtances both the Dorians and olians employed the T for the &,

as in the pronoun Tr, which they wrote TY. The poſleſlive, however,

derived from it, was written with either letter indifferently by the poets,

as ſuited beſt with their rythm and metre ; whence it is probable that this

variation was, in all inſtances, rather habitual than provincial. Both the

Engliſh and French now ſound the Tas an S before the vowel I in many

inſtances, particularly in the abſtract ſubſtantives derived from the Latin;

unleſs, indeed, that the Engliſh have now almoſt univerſally corrupted it 

$ into the barbarous ſound of the SH. The caſe is, that the L being only a ' T aſpirated. in a. particular manner, would naturally be confounded with

\ it in the different modes of pronunciation which habit or caprice give riſe

WAN to in languages not fixed by any eſtabliſhed rules of orthography, which 
 the Greek was not till the Macedonian conqueſt, when the later Attic be- 

came the common dialect; nor any of the modern languages till within this century, when the French and Engliſh made etymology their ſtand- ard, whilſt the Italians and Spaniards more wiſely adhered to pronuncia« 

—— <7 ü—ü—ᷓ— On -

(1) See VI. 1001, & ſeg. ed. Brunck.



ON THE GREEK ALPHABEI.0 tg tion; whence their words are ſpoken as they are written, and a foreigner,

who has learned the power of thei 5 knows ”-u to utter the ſounds _

which they repreſent.

The liquid conſonants are thoſe which ae of the nature both of

mutes and aſpirates, being pronounced by a ſuppreſſion of the breath in one part, and a conſtrained expiration in another, except indeed the R, which is uttered by the breath being violently forced between the tip of

the tongue and roof of the mouth, ſo as to cauſe a vibratory or jarring

motion of the former, by which the natural current of expiration is broken and interrupted (1). In modern orthography, the note of aſpiration. is al- ways affixed to it ; but this is not ſupported by the authority of any antient inſcription, though 1 it occurs in the common Roman form upon ſome of the moſt antient monuments of Grecian art now extant, ſuch as the coins of Leſbos, Tarentum, Croto, and Syracuſe. *

The Lacedzmonians employed this letter inſtead of the 2 in the termi» nations of their words, of which we have a curious example in the decree againſt Timotheus, the Milefian muſician, preſerved by Boethius in his Treatiſe upon Mufick (2), and more correctly re-publiſhed, from a Manu» ſcript-at Oxford, in the year 1777 (3); ſmall remains of it are alſo to be found in the Lyfiſtrate of Ariſtophanes (4) ; and Plutarch's Life of Pyr-

rhus (5); but the tranſcribers, not underſtanding theſe curious provincial peculiarities, have expunged them from the orators and hiſtorians, other -

wiſe we ſhould probably have had them in the other public acts of that people. This might poſſibly have been the caſe with the Z employed for the ©, of which, however, there is no trace in any written monument of the Laconians, though it occurs in the converſations attributed to them;

(i) A ture vu typrppuate; ( P) vn Qopas luihesvat· ra m TW Tpopuyy nv Oy Ty Tpaxa*

7 · MN

ri de ty Toig T01010de prjucrtiv, 6:09 KPOTEIN, OPATEIN.———T.% Morra & rr, %

puoay, jorge d ones Plat, in Orgy (2) Lib. I. c. 1.

(3) As this decreo i is a very important monument of 1 n conne ge with the ſubje & of this Eſſay, I ſhall conſider it apart at the end.

(4) Nanaicp for mana VI. 987, ed. Brunk, (5) Bk d n wapiorruv,' ous Marbprutec, wie, my gory Aug n bis ab en e d arme. ter Ko TE KAANGY GNNaps



wherefore, am rather inclined to think it a vicious habir of enten,

never authorized by orthography. To pronounce the L, which has an affinity with the R, the nine

motion abovementioned is ſtopped, and the tip of the tongue pteſſed againſt

the roof of the mouth, while the breath is forced out by the fide of it.

The N is pronounced by the tongue being entirely compreſled againſt the roof of the mouth ſo as to ſuppreſs expiration by that channel, which is, however, continued through the noſe.

The M has a near affinity with it, being equally anne "I a continu- ance of expiration through the noſe, whilſt it is in other reſpects r by a compreſſure of the lips.

Neither the form nor power of the naue have varied materially Gem what they were in the Latin alphabet, which is the oldeſt Greek except the Etruſcan. The Lambda has indeed been written ſometimes with the one and ſometimes with the other end upwards ; and the Latins retained one mode, and the Greeks the other, whence the two old forms were v and A, The Sigma was alſo repreſented ſometimes by a ſimple waved

line, thus 5, and ſometimes by one more complicated and angular, thus T,

which in the Venetian and Veletrian inſcriptions, and on the very antient coins of Sybaris and Paſidonia, is placed horizontally, thus M, while the

Mu is diſtinguiſhed from it by the angular lines being of deen lengths,

thus M, or more complicated, thus H. The laſt claſs of ſigns for ſounds are thoſe which enpreſent the different

tones of the voice, and which we, therefore, call vowels. Tones being iufinite in number, and varying in almoſt every individual,

the atranging them under diſtin heads, and reducing them to any fixed

and permanent rules, may be conſidered as the laſt refinement in language; a reſinement which the 6mple and determinate harmony of the Greek tongue ſeems to have been alone ſuſceptible of ; for none of the antient Oriental alphabets had any vowels (1) except the Phoenician, and that had

(1) The Shanſerit has; but whether that alphabet be original, like the language, I very much doubt, as both the forms and number of the letters ſeem to imply that it is made up from the ſpoils of others; and I belieye there are no very antient - inſcriptions to be found in it. The oldeſt that have been publiſhed are but little anterior to the Chriſtian æta.

2 . _ - rroperly

ON THE GREEK ALPHABET. . th properly only two, the Aleph and the Ain, ſignifying (as 1 ani inclited to think) merely the different degrees of aperture in the mouth, required to pronounce the words repreſented by the con ſonants (x).

The Greeks, even in the very earlieſt ſtage to which their Alphabet can be traced, had five; all which (except the Alpha, borrowed” from the Pheenicians) appear to be of theit own invention. The Latin, and-athet alphabets formed from the Greek, have confined thetaſc}yes to this num · ber, though wholly inadequate to expteſs the licentious variety of tones employed in the corrupt dialects of the moderns ; whetice they are obliged to repreſent many different ſounds by one letter, to the utter confuſion of all method and analogy in writing. The French, as I have heard from thoſe who have minutely ſtudied their language, protiounce the E only with more than ten different variations of tone; and in our own language the ſame vague and licentious utterance prevails, It the Greek, on the contrary, each vowel ſignified one tone, varied only in extenſion and ac- - cent; that is, in the length of time employed in the expiration of the breath, which formed it, and the degree of force and rapidity with which that breath was forced from the larynx. Vowels invatiably long are not properly diſtin letters, but, like the double conſonants, a fort of Sigla, by which the united ſounds of two letters were expreſſed by one mark. They are ſaid to have been invented by Simonides, and began tb be generally uſed about the time of the Perſian Invaſion, although the Athenians did not adopt them till the Archonſhip of Euelides, which was in the ſecond year of the xcivth Olympiad, 403 years before the Chriſtian #ra, and 77

after the retreat of Xerxes. The name of Gelo, King of Syracuſe, who 

died in the third year of the Lxxxvth Olympiad,-478 years before the: Chriſtian æra, is written, upon his Coius, with the Omega (2); and the Eta occurs upon the coins of the Rhegians, which, by the ſtyle of work- manſhip, ſeem to have been ſtruck nearly at the ſame period, or a very Fttle earlier. They have, however, the genitive plural written with the ſingle O (HIN ON, as thoſe of the Coians have, though ſtruck when the Omega was employed in the ſame word, as KQION, which we find upon

(+) Some, perhaps, will add the Jod; but, beſides that this letter was not properly a yowel, I have never been able to diſcover it upon any genuine monument of Phoenician writing.

(2) Sce Torremuzzi, Pl. XCVII. many


many medals of the iſland of Cos (1). Theſe double vowels muſt have relieved the Greek language from many ambiguities, eſpecially after the diſuſe of the ſimple aſpirates, which, on many occaſions, ſupplied their place, as I ſhall ſoon ſhew, Their having, however, been licentiouſly

uſed, ſometimes to ſignify the coaleſcence of two vowels into one, and

ſometimes the prolongation of a ſingle vowel by a ſucceeding pauſe, has

cauſed conſiderable confuſion both in the analogy and proſody of the Greek 

tongue, as 1 ſhall prove when I come to examine the metrical powers of the letters, to aſcertain which is the principal object of this Enquiry. It is generally ſuppoſed, that both the double vowels and diphthongs were unknown till many ages after Homer, as well as the double conſonants Z,

Z, and Y, which are in fact only abbreviated marks to expreſs two letters

by one character, like thoſe uſed in the manuſcripts of the middle ages, and copied by the firſt printers. This is, however, not quite ſo clear ; for the diphthongs are found in the moſt antient inſcriptions extant, though afterwards diſuſed, The firſt Sigean, written about ſix hundred years be- fore the Chriſtian æra, has EIMI ; but the ſecond, copied from it, pro- bably about fifty years after, has EMI (2). The Ol diphthong is alſo in the Veletrian inſcription, which 1s at leaſt as antient as the firſt Sigean (3). Upon a medal of Leſbos, more antient than either, we find the word NQIMYZHRQ, written from right to left, with the double vowels (4) ; but upon another, of a leſs antient ſtyle, the word NOIAT AA has the ſingle O in the genitive plural (5). The firft of theſe words ſeems to be a myltic title belonging to ſome obſolete dialect, for it has no apparent affinity with the known roots of the Greek language; but the laſt is probably the geni- tive plural of FETHE, employed equally as a myſtic title. Words equally anomalous with the firſt occur upon the very antient medals of Side, in Pamphylia, 'written alſo with the double vowels ; but, as ſome unknown

(1) See Dutens, PL. IV. Fig. 41 Pellerin, Pl. CI. Fig 1; and 8 "Wong Pl. XXIII. Fig. 2. Similar medals are in the cabinet of the Author.

(a) See Pl. II.

(3) See Pl. I. Fig. 2»

(4) See Pl. I. Fig. 4.

(5) The medal I ſaw in the cabinet of the King of Pos There is one exactly ſimilar in the Hunter collection, except the letters, See Comb, Pl. XXXIII. Fig. 3.

e ee

a y E - . = ” 2 = py N * 8 1 3 Ce. D „ < Y * q — 44 2 een 4. W FP — I . q 4 # * 1 V l — A p » os * =_ "4% * 1 . — ">: * - 19 I L

a "IF - $7 ad n = PS 2

Ks " 2 % < l l F — rn * £ „ S

bs REPS N **

  • 8 en


characters are mixed with them, it is impoſſible to decide whence they came, or to what language they belong (1). It is equally impoffible to aſcertain the date of theſe antient medals ; though we may ſafely pro- nounce them to be as old as any written monuments extant, except the hieroglyphical inſcriptions'of Egypt; and, perhaps, ſome Etruſcan or Pe- laſgian antiquities. Thoſe of Leſbos, in particular, are of the moſt antient fabrick known; and, from the numbers in the ſame ſtyle which have been found, muſt have been ſtruck when that pres gay om FIRES and power. . This, according to ahi cormpurtion of Ruſebius, was daring the ſeventh century before the Chriſtian æra, the Leſbians having poſſeſſed the empire of the Mediterranean from about the xxvith to the xL111d ' Olympiad; © whereas Simonides did not flouriſh till the end of the Lxth Olympiad, full ſeventy years after (2). I am, therefore, perſuaded that the double vowels were uſed in Aſia before the time of that poet, their ſuppoſed inventor; . who might, nevertheleſs, have brought them into Greece, and rendered the uſe of them more popular and general. The age of Homer is, how» ever, ſo much anterior to all monuments of art, or authentic records of hiſtory, that we cannot even tell whether or not he had the knowledge of any letters ; there being but one paſſage in his Works where writing is mentioned, and that is ſo equivocal; that it may mean either ſymbolical h or alphabetical writing (3). _ - 1 The form of the double vowels ſeems not to have varied conſiderably till the age of Hadrian, when the{Omega, which was before written QorQ, - was, as Euſtathius obſerves, made out of two upſilons, and written (like our W) . 1 attribute the introduction of this form to that period, be- cauſe the Egyptian medal of Antinous is the oldeſt monument of art, of which the date can be aſcertained, that exhibits it; other medals of the

ſame perſonage having it in the antient form. As to what Abbé Winkel- 

mann ſays, of its being upon the medals of the Macedonian kings of

(1) See Pl. I. Fig. f, from a medal in the cabinet of the Author. (2) See Euſeb. Chron. lib. II. verſ. J. Hieronym.. -'

egen J % onpals A

(3) —

Teadag i wivaw wrvxry Wujoploga worns b *

Aufas 9 'nwyn 7 w1/01gw, of a rohre. Il, Z. 168. D 2


Syria (1), 1 can take upon me to aſſert that it is untrue, no ſuch medal having yet been diſcovered either of the Syrian or any other of the Mace- doniau Dynaſties ; though it is probable that the learned Antiquary was deceived by ſome counterfeit, he having no knowledge of coins (2). It ia, indeed, upon a braſs vaſe, preſeryed at Rome, which appears, by the inſcription, to have been preſented by King Mithridates to a Gymuaſium ; but this Mithridates was probably the petty prince of Thrace, who reigned in the times of Trajan and Hadrian, and not the great King of Pontus, whoſe taſte and magnificence would ſcarcely have condeſcended to make fo paltry a preſent, and much leſs to have put his name upon it. The ſame kind of Omega is, indeed, in the names of the two artiſts, which are in- ſcribed upon the two celebrated ftatues of the Hercules Farneſe, and the Torſo of the Belvidere ; but as theſe artiſts are not mentioned by any an- tient writer, it is probable that they lived under Hadrian and the Anto- nines, and that the ſtatues are copies from more antient works. The Co- loſſal head of Antinous, in the villa of Mondragone, at Freſcati, and the buſt of Trajan, in the collection of Mr. Townley, prove that there were then artiſts capable of the executive part of either of theſe figures, though the grand ſtyle of compoſition, which peculiarly diſtinguiſhes the laſt, had been long extinct. It is probably a copy of ſome well-known groupe of Hercules ſtrangling the lion, the attitude appearing to have been nearly the ſame as that in which he is repreſented upon ſome of the {mall ſilyver coins of Heraclea, in Sicily (3). The proper mode of pronouncing the Greek * has dan a ſubject of much controverſy ever ſince the revival of learning in the Weſt; it ' having been ſoon diſcovered that the Byzantine Greeks, the only teachers of the language, had long loſt the art of ſpeaking it, though they conti- nued to write it with purity, and even elegance. Eraſmus firſt compoſed- a whimſical dialogue upon the ſubject; and ſoon after Cheke, Profeſſor of Greek in the Univerſity of Cambridge, undertook to examine it ; but his work was anticipated by an ond, publiſhed in the year 1542 by Stephen

509 Hiſt. des Arts, lib. IV. p. 122.

(2) See Hiſt, des Arts, tom. III. p. 931 where he has dabndbes one of the moſt 7 ling modern counterfeits ever executed, as a true medal of Antigonus, King'of Aſia.

(3) See Ter Pl. XXXV. 4 & 5.



Gardener, Biſhop of Wincheſter, and Chancellor of the Univerſity, ſtrietly commanding that the mode of pronunciation eſtabliſned by the modern Greeks ſhould: be continued; by which the vowels H, 1, and 1, were cons fidered merely as different ſigns for one ſound, the diphthongs Ol and ET for another, and Al and E for another. Cheke and his friends found no difficulty in confuting theſe abſurdities; but neither he, nor thoſe who have followed him in the enquiry, have afforded us much real information, except that which was before given by Dionyſius of Halicarnaſſus. 4 The A,“ ſays that Critick, © when extended, is the moſt ſonorous of the « Jong vowels. It is pronounced by the mouth being very much opened, e and the breath forced upwards, Next is the long E; to pronounce « which the mouth is moderately opened, and the ſound, following the breath, preſſed down about the root of the tongue. Then comes the long O, which requires the mouth to be cireular, and the lips contrafted round, againſt the cutward edges of which the breath muſt be ſtrongly « impelled. The T is lefs ſonotous ; for, the breath being conſtrained by

& a conſiderable contraction of the lips, the ſound produced is flender. In- « ferior to all is the 1; for, the mbyth being but little opened, there is a « collifion of the breath with the teeth, and the lips are not employed in « elevating the ſound (1).” This paſſage entirely ſubverts the authority of the Byzantine Greeks, as well as that of our own ſchools, none of which teach the true pronunciation of the vowels, except perhaps the Scotch. The Critick has conſidered the long ones rather than the ſhort ones, not becauſe there was any difference in the mode of pronouncing them, but becauſe tone can be better illuſtrated and aſcertained in a long ſound than a ſhort one. It appears, from what he ſays, that the A was pronounced as the Italians now pronounce it, or as we -pronounce it in the words VAST, PAST» &c, The E was alſo as the Italians now pronounce it, or

(1) Abris N Tar paxgiy * T3 , rar Iurilnras* 3 yap auer, Th claare; in} ben 1d e d Proouire wpes Thy Sparir. Hibriger N vd v. Gre xndrw wie Tw Baow Tis yuoon; hgiidn vd 1x9» AN, 62 wn &vw, 35 pimples aroryopiry. Tehror N 73 „ reoyylaural vu yag b airy d rina, wy Sir 7% x, The T1 amy v Trina wil vd Gxgorbuior i. ifs N rh Thru rd v* ie ne add 7a Xian eures yropirg AE enbyrras, x, ries ie © I xe · Taxorror A S 76. ige Thy Gibrrag TV yag 4 ngornou, vi S ν yinurar, winger en 18 care, & tx vTIMpTrguverruy e roy zx. Dionyſ. Halicarnaſſ. ws curbio.

9 — , as . 3 * - * *


as we pronounce the A when followed by a conſonant and mute vowel, as in the words MATE, PLATE, .&c, The Italians have alſo the true pro- nunciation of the O, which we have miſerably corrupted, except when followed by a conſonant and mute vowel, as in the words MoD, BoDE, &c.

 As for the T, 1 am in doubt whether any modern nation pronounces it 

[ exactly as the Greeks did: the Italians follow the Latins, whoſe U corre-

\ 1 ſponded to the OY diphthong of the Greeks, the true pronunciation of

which is retained by the French in their own ov. We pronounce it as 
the diphthong EY in ſome inſtances (as in raab), and in others, as the 

1 French pronounce the ſame diphthong (as in Tr), a barbarous ſound

"I unknown to antiquity, Perhaps the neareſt letter to it in modern alpha- -

Wh SED bets is the French accented U; the ſound of which is, indeed, poor and

' lender; but ſuch Dionyſus informs us that of the Greek T was. WY

  The vowels have varied but little in their forms, except that the Upſilon 

was antiently written like the Latin V, and the Iota by an indented line,

thus 5, to diſtinguiſh it from the Gamma, which was repreſented by the

ſtrait perpendicular line. The confuſion between theſe two forms probably

produced the I conſonant; which ſeems, in the Roman alphabet, to have

had that affinity with the G which it ſtill retains in moſt modern languages.

7 27 ef 5 - 4: 3 6

q SW

  • Y

. =

. o 

1 _ « ; [ 9 "=

  • -

L * 1 = * q .

A 1 0 t 1 K


n .

- >; - g rg of Te

l us as 5 lt e * * : Zee



s k G T ION ir.

AVING thus conſidered the letters as notes of articulation, aſpira-

tion, and tone, it remains to be conſidered in what modes aud de- grees particular acts of vocal utterance were lengthened or ſhortened, in proportion to the number and claſs of the letters employed in repreſenting them; for, as the Greek Alphabet was adapted to the language, and not the language to the Alphabet, we ſhall find the practice perfectly accord with the theory, unleſs where local or vicious habits corrupted it. Even there we have the peculiar advantage in this language of poſſeſſing the Works of a poet (the moſt elegant, correct, aud perfect, of all poets), who lived before many ſuch habits had been formed, and whoſe writings, there- fore, though defaced by the varniſhes of criticks, grammarians, and tran- ſcribers, are compoſed of materials ſo pure and ſimple, and executed with ſuch preciſion and regularity, that we can ſtill trace the minuteſt touches of the maſter's hand, and aſcertain, with almoſt mathematical certainty,

the principles upon which he wrought (1). For this reaſon I ſhall admit

(1) This character of Homer's poems may, perhaps, ſtartle thoſe who are accuſtomed to receive their opinions, ready- formed, from the futile, but pompous, aſſertions of certain ſelf-created judges of literature; whoſe deciſions, to the diſgrace of on age, are not un-

popular. One of theſe has lately pronounced, with all the technical jargon of a profeſſed book- maker, that the Greeks had no ears for metrical harmony ; but that all their poets, and more eſpecially Homer, continually tranſgreſs the rules of their own proſody ; their verſifi- cation being, as he ſays, always . . and generally rough and unmuſical, and termi- nating in what he calls cacophonies, (Recherches ſuc les Grècs).

That there ſhould be a mind ſo perverſely organized as to form ſuch opinions as theſe, when nurtured in the pride of pedantic ignorance, I am not at all ſurprized, for I have ob- ſerved as many luſus naturæ in morals as in phyſicks; but that there ſhould exiſt one, ca- pable of forming or comprehending a ſingle ſyllogiſm, and yet ſo deſtitute of common judgement and diſcretion as to publiſh ſuch paradoxes to the world, and thus become the herald of its own imbecility and deformity, is ſcarcely to be accounted for, even in the wide extent of human inconſiſtencies. .



no general rule or principle of metrical quantity that is not juſtified by the practice of Homer; having found that his practice is always founded upon reaſon and analogy, whereas that of later poets was often regulated by lo- cal and temporary habit.

Upon his practice, therefore, and the . before ſtated, I venture to draw the following general concluſions: /

I. A ſingle vowel, repreſenting a ſingle act of vocal utterance or expi-  

141 ration, muſt neceſſarily be ſhort, unleſs lengthened by a ſucceeding pauſe / 41H or obſtruction of utterance ; for the proper definition of a ſhort ſyllable is, one that occupies only the time uſually allowed to a ſingle act of vocal ut- terance;z whereas a long one is that which occupies the time uſually ap- propriated to two, either by being really a coaleſcence of two, or elſe by being delayed or impeded by ſome adſcititious pauſe or obſtrution.

If there be no ſuch pauſe or obſtruction, and the ſucceeding word be- gius with a vowel, this vowel, if ſtanding alone, or terminating a word, WI; will be ſwallowed up, or, as the grammarians ſay, elided ; for tones, un-

— leſs divided by a pauſe or ſuſpenſion of the breath, naturally coaleſce, or nl flow into each other,

"5 ONS The Greeks, however, in their Heroic or Hexameter verſe, admitted of 
il an arbitrary or artificial pauſe, and often ſuſtained one vowel entire before 

WI another in a different word; but in dramatic poetry this was not allowed;

 / bs PONY Li neither did the Latins, in their Heroic verſe, admit of ir, otherwiſe than  

1 WA pe m—— fl as a licence, juſtifiable by the example of the Greeks, when Greek words © 5 24. dale. Gel were employed. I cannot indeed but think that it crept originally as a li- Ar 2 _ 4 cence, introduced by the loſs of the aſpirates, into the Greek language; 75 Un fee Py od and that it was never really juſtified by the practice of the antient poets, 17 whoſe works, according to the preſeut orthography, afford ſo many in- 10 a ſtances of it; for, if we reſtore the aſpirates according to etymology and 0 antient practice, we ſhall find ſcarcely any inſtances in Homer that may 4 not be cured by a ſlight change in the order of the words, in which the .

ll Manuſcripts and old editions continually differ; or the inſertion of a par- 
ef , ticle, always admiſſible, and often required by the ſenſe. In the genuine 
2 (our bes) 55 



7 1 j

= 1 _

4 b em ö Heſiod too, I know of only four inſtances, except thoſe where IN the aſpirates are wanting; and of theſe four the emendations appear ſo ob-

  1. C:

u ; Wy —


vious, that I ſhall venture to propoſe them, tough without any better aus

1 than my own conjectures.

For agwge: n, J would read afage v A 1).

For Saas ba- 46 v avoror, wabanrirle Y (i. e. 700 uit egi / Fre aN ſhould be N axacr (2) or perhaps only have the paragagie N added to the firſt word ; for the Bæotiaus frequently dropt the apicates, as appears, from. the very antient medals of Thebes, upon, which *. of that city is written with the T ipſtead of the © (3% .

Meooy ede may be pweron” eniduine (4) ; the adjective luinvuag or MET ZE FOR occurring in the feminine, . psgox, in VI. 767, according to a Manuſcript collated by Grævius; and this reading was preferred both by

him and Robinſon (I think rightly), oneieh ſanding the FR Manner

in which M. Brunk has rejected it.

Later poets, however, have continyally 1 of vowels ſuſtained before other vowels in different words ; but theſe poets may be conſidered as writing in a dead language; for. ſuch the language employed in Heroic verſe then was; both the words and flexions being taken upon the autho- rity of the antient and popular bards, when no longer known as the ordi- nary means of ſocial intereourſe. Had they, indeed, poſſeſſed the works of thoſe antient bards in their genuine ſtate, their imitations of them would at leaſt have been exact, as thoſe of Vida are of Virgil; but between the age of Homer and Heſiod, and that of Apollonius Rhodius and Theo- critus, the alphabet, orthography, and pronunciation of the Greeks appear to have been greatly altered; and with them, of courſe, the laws of pro- ſody, which regulated the old Hexameter verſe. Hence, in that verſe, there is an appearance of arbitrary licence in the extenſion and abbreviation of the ſyllables, which none of the metres that employed only living dia · lects admitted, and bin cannot believe to have n e PL fn

(1) VI. 318, ed.

(2) VI. 685, ed. Brunk-

(3) Theſe medals are very ſcarce. I do not recolle& to have ſeen more than one, which is a tridrachm, with the vaſe on one fide,” and p erg rw on ws oft; 217 Mo in _

cabinet of Mr, Vandamme, at A. C4

(4) VI 755+ 1 , by 0 . 47 E¾ũ when


when all the words and flexions which it employed were in familiar uſe, as they undoubtedly were in the time of Homer. The Iota ſubſeriptum of the dative caſe being, as will be more fully ſhewn hereafter, a vowel of itſelf, regularly affixed to the preceeding one, as it is in other declenſions to the preceeding confonant, that preceeding vowel 1s guarded by it, and therefore not neceſſarily elided. Hence the 13 terminations in y and ꝙ often remain long before another vowel, The » 9 is alſo ſometimes long before another vowel when affixed to an aſpirate, iN as in the third perſon ſingular of the Aoriſt and paſſive, abb, the — 14 reaſon of which will appear when we conſider the metrical power of the wy aſpirates. ll IL A ſingle vowel before a ſingle mute conſonant muſt neceſſarily be ll ſhort, unleſs there be a pauſe between them; for, as the conſonant termi-

 nates the ſound without adding to it, there will of courſe no pauſe accom 

pany it.

; If, however, a ſecond mute conſonant follow, either in the fame or a 

different word, the ſyllable, though not the vowel, will neceſſarily be I long; for, as its concluding conſonant fignified a ſuppreſſion of the breath

 which produced it, and the opening conſonant of the next ſyllable the 

1 commencement of a new act of utterance from an equal ſuppreſſion of the

breath, there muſt neceſſarily be an intermediate act of ſuſpenſion or expi- 

f ration ; which, how ſhort ſoever it be, will require a pauſe ſufficient to [ lengthen the firſt ſyllable, to which it muſt neceſſarily be added; becauſe

 + A the ſecond only began with its opening conſonant, 

I I know that this intermediate act of ſuſpenſion or expiration, which t conſtitutes the pauſe, is ſeldom perceptible in modern pronunciation, eſpe-

cially that of the Engliſh, who never utter two conſonants of the fame 

Will *N organs diſtinctly when they come together in the ſame word. The ſecond i D and T, in the words Abpzp and eiTTED, are never uttered, but only 4 ſerve to give the firſt more than common force and emphaſis. In the ſame "nr manner we pronounce the Latin words QyIDDAM, QUICQUAM, &c. and the few Greek which there are of this deſcription. (118 The K being, as was before obſerved, formed out 'of two ſingle conſo- 1 nants, is frequently employed by Homer with the power of a double one ; 14 but,

  • * "x .

n ra. TY WL TE Fog AI . , = _ «A * 12 K . r SE or 4 1 340 Cx" N 1 *w 2 Nr „ TGT =

2 8

as a . <a 22 7 1 4 of * 4 de 8 0 44 a 2 1 2 3 * 8 _ PEO bas RY" 40 7 A ey 48 * Wr 2 k 8 . IT. > "46. 02.4 Ks e r — 0 " - 8 "II GC wt ads... 6 Sg 1 . L 4 T (4 ; N „ en. mn 3 - r 2 ? Wa — = * „ e „ e mee R ‚ ‚ N Ea. VS EF * * »# 2 — *

7 b 8

Sv x —_

— 4 3 > 1 * * ve „ * — * bt A, . y

  • = * We TIT — RO VET F 8 & A

. 9 3% ” -



but, when we find the ſame power given to any other mute conſonant, we may en that it was originally aſpirated, or that ſome letter has been omitted. TTaE TT is uſually derived from TTT Ox, little, which might account for the extenſion of the firſt ſyllable, if the ety mology was admiſſible, . which, I think, it is not; for, though the Greek names were all deſcriptive titles, they were never titles of diminution, or degradation. The aſpirate, we know, was dropt from the T in the dialect of the antient . 7 5

Thebans (1); who, therefore, wrote the name of their city TEBE, inſtead of @HBH. The ſame. pronunciation and orthography probably prevailed among the antient Ætolians, who muſt conſequently have pronounced and written the verb dus or @YFQ, TYFQ ; and if the name Tudeuę be derived from it, as it appears to be, it muſt of courſe have been written oo the ſame plan, TYFAEFE.

The orthography, indeed, of moſt of Home s names may be conſidered as merely traditional, for the oldeſt inſcriptions, in which any of them are recorded, are of an age long poſterior to his; and no reliance is to be placed in the copies of antient authors which have come down to us; for even a name ſo well known, and of ſo late a date, as that of the great King of Pontus, has not eſeaped corruption: upon his coins, as well as upon every other antient monument that bears his name, it is uniformly written MIOPAAATHE, according to its ety mology from Mlep ATL; but in

all books, both Greek and Latin, it is as uniformly MI@GPIAATHE. _

Whenever, therefore, this tradition is oppoſed by radical etymology, or

metrical analogy, we may, I think, venture to pronounce it wrong. ABTAOZE was probably written with the Digamma, ABTFAOE, it being derived from the verb BTF. MAETIZ-ITOE was MAZ TINTE-IN OR, 4-40 in the ſame manner as OOPMINTZ-IN TON, EAANINDE-INDOE, and J. a 9 FAY .f other words of the ſame claſs; ſome of which have dropped, and others retained the N (2). In ſome editions of Homer we have, indeed, wagiyi, in II. Y. 500 ; but the true word here, as the Venetian Scholiaſt has ob-

(1) As in the medal before cited,

(2) In Heſychius we have (3; pagryt. Heinſius, ng, 3 the F in the laſt word to be inſerted erroneouſly ; but it is in reality the antient form with the uſual variation, Marrirz for Marrmrz, the ſame as mhagrys for TIAALTINTE, a word of ſimilar meaning, and formed upon the ſame plan, from a different yerb.

E 2 ſerved,

4 2 — — —— — —— —— - = —_ —-

m ä —̃— 7"

EMMEN, EMMEZENQI, NPNITTOZ, &c. for 1 Eu, ty pita, Tpory yvos, '&c,


ſetved, is nag, or rather MALTII, the Jonic dative of an obſolete word, MAETIE; with the A elided in the ſame manner as in herr or GET I for SE TIal.

Such elifions are extremely common in the Greek language, as muſt be obvious to every one who has even curſorily examined it. The omiſſion of the N before the palatial conſonant is general in the preſent ottho-

graphy, though its place is uſually filled by doubling ' the conſonant, or

adding another of the ſame organ, as in , a, &c.; which appear, from antient medals and inferiprions, to have been originally written, as they are ſtill pronounced, ENXOZ, ANKAQN, &c.(1) The L was ſtill more frequently elided, as being a letter the ſound of which was abhorred by the reſined ears of the Greeks; whence great con fuſion has been intro»

duced into the tenſes of the verbs, as I ſhall more. particularly obſerve -


The firſt fyllable of ſome words . of the n is 09770 are occaſionally pronounced long, though conſiſting only of a ſingle vowel followed by a ſingle mute conſonant, as Amonauves and arvorieoda, in which the II was, by ſome provincial habit, pronounced double; or (what is more probable) delayed in the utterance by the muſical pauſe or exſure 3 for this licence never takes place but in the firſt ſyllable of the foot; and, as all very antient verſe was ſung to the lyre, there might have been ſome particular ſtreſs or pauſe in the accompaniment on theſe occaſions.

Ilapberonimrys ſhould probably be APOENOTITIITHE, it being derived from wagherog or o, and the omiſſion of a letter being marked by the circumflex. The letter might, however, have been elided in the time of Homer, and the cuſtom of pronouncing the ſyllable long continued after

the change in the orthography. A ſingle vowel is often long before the adverbs z and dee; but the

firſt of them is ſometimes written $4»; and, as the ſecond is derived from it, we may conclude that it was written in the ſame manner, which is in-

(1) See Torremuzzi, Pl. XLV. Fig. 9 & 10; and Comb. Pl. V. Fig. 2. In the He-

raclean inſcriptions, however, which are in the Doric diale&, and about 300 years before

the Chriſtlan era, the N is more conſtantly changed than at preſent, as appears from


Wn ” s ”



4 £


2 3 by 7 J 4 2 4:

in. x

i. * . „


deed more conformable to etymolagy, the toat being F or TEINQ- Written in this mode, the metrical anglagy becomes peifectiy regular; for,

III. A fingle vowel followed by an aſpirate or liquid, either in the ſame or a different ſyllable, or even preceeded by one in the ſame ſyllable, may ; be either long or thort, ſince the conſtrained expiration, employed in ſounding the aſpirate or liquid, is a captinuation of the vowel ſound differ- ently modified by the approximation gr compreſſion of the organs of ſpecch, and may therefore be ſhortened or lengthened arbitrarily, according as the conſtrained expiration is continued . for a greater or leſs time. Heuce both the aſpirates and liquids ate often ritten dauble when etymology requires that they ſhould be ſingle, as in , eh, $0,067.41, KC. * is no more reaſon for Writing the lettets double than in 7 aopog, & anĩvigeucto, Gini, & c. pronounced q Ne Je, amerugerto, GH, We Ariſtarchus appears to have diſapproved of this departure from etymology (1), which certainly ought to be entirely adhered. to, or entirely neglected, for pronunciation; as the preſent orthography, beiug regulated upon no prin- ciple, gives the appearance of anomalies where thete are none. In the flexions of the verbs the doubling the T is, however, ſometimes regular; the old Aolian and Dorian terminations in -Zan forming the future in Tak xa, contracted to - TAS, and thence, hy the &'s being elided for the ſake of ſmoothneſs, to · xx. Neyertheleſs, it appears from the He- raclean Tables, the moſt complete and perfect monuments of the kind ex- tant, that the antients adhered more to pronunciation than etymology, whence, in addition to other local peculiarities, we have uniformly en, ET TON TAI, ExZHT ALI, &c.

When two aſpirates or liquids come together, or one of them be joined. to a mute. conſonant, this conſtrained expiration will naturally be length- ened or obſtructed, either of which will prolong the ſyllable. Neverthe- theleſs, the Attic writers, whoſe dialect was ſpoken more cloſely and ra-

(1) See Schol. Ven. in Il. K. 258, See alſo Erneſt. Not. ad II. M. 281; and Calli- mach. Hymn. in Del. 110; and Clarke ad II. N. Vſ. 1; where, after having very ingeni- ouſly and pertinaciouſly defended an erroneous opinion throyghout his firſt volume, he very effect ually, though not very openly, recants it; and thus at once ſubverts the fine - drawn

ſyſtem of metrical quantities, which he had laboured to eſtabliſh through all ** preceeding notes.


1 TV


AN ANALYTICAL ESSAY pidly than that of the antient Æolians and Ionians, pronounced the vowel

ſhort before TM, KN, n, and TM; but, in all inſtances of this kind, the E, k, IN, and T, merely mark the commencement of utterance, or

preparatory ſuppreſſion of the breath, and were therefore very ſlightly, if

'at all, pronounced in the rapid and' conciſe ſpeech of the Attics. In the

works of Homer, however, the ſyllable is, in ſuch caſes, always long (1), though he admits of the A or P to be joined to a mute conſonant without

extending the preceeding vowel; for both theſe letters expreſs tone as

much as articulation, and therefore are properly called ſemivowels. Ac- cording to Dr. Clarke, indeed, he makes the ſecond ſyllable of Aqur rig ſhort ; but as it is always long in the ſubſtantive A:»wT7%, I believe that

acute critick to have been miſtaken, and that we ought to pronounce the

laſt vowels, even in the oblique caſes, as one ſyllable; the I having no other power in this, as well as many other inſtances, than the Y in our

words YEAR, YAWN, &c. in which it is rather an aſpirate than a vowel.

The firſt ſyllable of requrs is, nevertheleſs, ſhort in our preſent copies in one inſtance (2); but the Harleian Manuſcript, collated by T. Bentley, has t, which is probably right, though the vowel might poſſibly be ſhort, even according to the old Ionic pronunciation, before two liquids of the ſame organ; and if Homer had any other inſtance of it, I ſhould prefer the common reading ; but when a general concluſion is drawn from ſuch

(1) Hence we may conclude that the Batrochomyomachia is not Homer's, but a bur- leſque imitation of his manner by ſome anticnt Attic poet, who, though he adopted the words and expreſſions of the old bard, formed his metre according to the pronunciation of his own country.

With equal confidence we may pronounce the Margites to have been a forgery, though there are only four lines of it extant, and three of thoſe are quoted as authentic by Plato and Ariſtotle; but in theſe we have a compound verb with the augment- upon the prepoſition

(vricare) ; which Homer's grammar did not admit.

Similar objections may be made againſt the hymns and epigrams, ſome of which have been ſtated by Clarke, and others will be noticed in this Eſſay.

Theſe peculiarities are more certain proofs of the authenticity of the Iliad and Odyſſey than any hiſtorical evidence would have been, for they ſhew that the moſt antient imitators and moſt learned readers of thoſe poems never obſerved the diſtinctive ſingularity of their

dition, and therefore could neither have forged or reverſified them, as ſome have ſuſpecled. (2) II. x. 707.

a num»


a number of examples as neceſſarily occur in the two long poems of the Iliad and Odyſſey, a ſingle exception is of courſe ſuſpicious.

We have alſo in out preſent copies one inſtance of ryraſaih e, U), t two of aròpo rug (2), and ſeveral of avdpsiporry; (3), with the firſt ſyllables ſhort, But the firſt word is, in the Venetian Manuſcript, written very properly

exerpubes 3 and the ſecond, as Damm has obſerved, ſhould be adh, as it is preſerved in a citation by Plutarch (4). Aldus's firſt edition of Plutarch

has, indeed, w»poryra; whence Erneſti ſuſpects that 4poryra is only a conjectural emendation of ſucceeding editors (5). Whether, however, it be ſo or not, it is certainly the true reading; for, beſides the analogy of metre, ſupported by the uniform concurrence of ſuch a number of in-

ſtances in various dialects, the very principles of the language do not allow ſuch a word as av3gorns to exiſt, any more than thoſe of our own tongue

ſuch a one as manneſs ; for the Greek abſtract ſubſtantives in · ru, like ours in -ne/5, are all neceſſarily derived from adjectives, and not immediately

from other ſubſtantives: ae, therefore, being the adjective ſignifying manly, ey3pworns muſt have been the form of the abſtract ſubſtantive ſigni-

fying manlineſi, if any ſuch in this claſs had ever been formed, which I do

not find that there ever was. Even if there had, it could not have been known to Homer; for the adjective does not appear to have exiſted in his

time, and prior to that the abſtract ſubſtantive could not have exiſted, any more than, in our own language, the ſubſtantive manlineſi could have pre»

ceeded the adjective manly.

Avdperporrns occurs only i in the dative caſe as an epithet to Mars G

avJeuperry), and, as it is now read, has the two firſt ſyllables ſhort, to the utter ſubverſion of all metrical analogy. The Leipſic Manuſcript has an- dpuporTy, which 1s little leſs objectionable, unleſs we elide the A, as in avsp, aud write evan avg pry, or, In antient letters, ENETAAIQI (6) ANPI-


(1) II. a. 274. 0

(2) II. x 363 & n. 6,

(3) I. 3. 6514 Hz 166; ©. 264; Pp. 259.

(4) De Poet. audiend.

(5) Ad Il. n. 857.

(6) Thus is this title written on the Maſtrilli vaſe, found at Bari, in Tt and pub» liſhed by Mazochi, which I believe to be right, for Suidas mentions the exclamation En;



— N "vo


32 AN ANALYTICAL ESSAY OTR. We may, indeed, ſuppoſe the preſent reading to have been

pronounced ENETAAT!' ANAPEFSONTHT, though the QI or © of the da-

tive caſe is not often elided. I wiſh there was any authority to write API- or EPIGONTHI, which, I think, would improve both the ſenſe and metre, and which I cannot but ſuſpe& to have been the original word, though it dves not now occur any where, It is, however, equally conſiſtent with the idiom of the language as epCpepurys, #pmudy;, eue, &e. 3 and its bo- ing little uſed was the natural cauſe of its being corrupted,

When a conſonant aſpirate follows a mute, as in the Z, E, and v, the preceeding vowel, or rather the ſyllable taken eollectively, muſt neceſſarily be long; for though the A or TE, the For KZ, and the B or IE, are each ſigniſied by one character, they never completely coaleſce in ſound, there being neceſſarily a pauſe, however ſhort, between the ſuppreſſion of the breath, which produces the mute conſonant, and the conſtrained 11 ration, which produces the hiſſing aſpirate (1).

Bat when the conſtrained expiration preceeds the entire ſuppreſſion, it

ſeems only a preparatoty or inttoductory part of it; for the conſtraint is it-

ſelf / .cdmmplete ſuppreſſion, which a continued approximation of the or- gans of the mouth to each other would render complete as ſoon as they catne into contact. The T, therefore, ſignifying the act of approximation, and the mute conſonant, which follows it, that of contact, both are only different ſtages or gradations of one exertion, and therefore form, when thus united, only one diſtiuct articulation; which may, nevertheleſs, be contracted or extended in the utterance, according as the idiom of the lan-

and the Latin verbs EN ECO and vnc, and the Greek ſubſtantive K TZ, are apparently derived from the ſame root, written according to different dialects xa, . ENE Ta, and EN.

(1) I am aware that there are ſome very learned perſons who have been of opinion that the ⁊ preceeded the mute in forming the double conſonants ; but I think, if this had been the caſe, the Doric verbs terminating in ran would have been written like the Attic and Ionic with the Z, otherwiſe the difference would have been to the eyes, and not to the ears, by which we know, nevertheleſs, that all the variations of diale& in the Greek lan- guage were perceptible, The K alſo inſtead of the ꝝx would have been elided before a conſo- nant in the prepoſition Ex or EK, and the Latins would have written ſuch Greek words as



= YT FY ed


guage, or cuſtom of the country, require. The extenſion, however, When

it takes place, will not be in the ſyllable in which theſe letters are em -

ployed, but in the preceeding one; for the delay cauſed by the hiſſing ſound is not an extenſion, but a ſuſpenſion, of utterance, w. ich utterance only commences with the conſonant that immediately preceeds the vowel.

The Greeks, in almoſt all caſes, admitted this ſuſpenſion, ſo as to make a ſhort vowel, preceeding a L and mute conſonant, long; but the Latins, whoſe language was leſs flowing and melodious, and ſpoken with more abbreviation and rapidity (1), often Py it over, ſo that ma your in many inſtances remains ſhort,

Homer, however, has i5/a:#, which * would Gd bs l

on the coins the name is always written with the Z, which might never- theleſs have been elided in the earlier dialects, for I know of no coins of this city which do not appear evidently to have been ſtruck after the Pelo- ponneſian war. He alſo makes the vowel ſhort before the names Zaxuibog and Zi, which ſome would therefore write Laxurbog and Ta but it is more probable that, in the old Ionic dialect, they were written AAKTN= S0 and AEAEIA, like the Zanclean medals, which were ſtruck by one

of the moſt antient Ionian colonies, and which have uniformly AANKAE

for ZANKAH (2). For the ſame reaſon the vowel is ſhort before the word Exapoay3pg, which was antiently written Kah bös, as it ſtill is in ſome manuſcripts and old editions. In one inſtance we have alſo mapa gabuw; but in Euſtathius it is, more correctly, wap gabuy (3)

It was either from not conſidering this, or, more probably, from being . Nartled at an apparent irregularity of grammar, that Ariſtarchus fo injudi- ciouſly changed the antient verſe, which deſcribed the ſcene of action be-

tween the Greeks and Trojans, from peooeyvg worapon Ka, , xa gojpcs 

Nn, tO preronyu; Eippcerrog, 108 Barlow powur (4) ; which, being preſerved in our preſent copies, has effectually puzzled the geographers who have attempted to fix the ſituation of Troy; for there is a chain of mountains between the ſea and the conflux of tho rivers which the Greeks do not *

(1) Plutarch. in Demoſſh. init. 4;

(2) See Torremuzzi, Combe, &c. as before cited. 3 (3) Od. x. 327. (4) II. z. 4. e

  1. % ,.- -


pear ever to have paſſed ; and in the XXIſt Iliad the ſcene is evidently be- low the conflux, otherwiſe the Scamander could not properly call upon the Simois to aſſiſt him in drowning Achilles. According to the old reading every thing is clear, the ſcite of the city being about the village of Borna- baſchi, where are ſtill the ſprings deſcribed by Homer ; which, flowing down into the plain, formed a lake, ſtill viſible, between the outlet of which and the river Scamander was the field of battle (1); This outlet was probably once into the Scamander, whence the fountains are called wy Exapayſp, They were two in Homer's time, one warm and the other cold: but Mr. Wood ſpeaks of only one; and the ſubterraneous channels may, perhaps, now be joined by the earthquakes that have fre- quently altered the face of that country,

By not duly conſidering the power of the aſpirates and liquids, ſome of the moſt acute and learned Criticks have embarraſſed themſelves with Imaginary difficulties; and then, by endeavouring to remove them, raiſed real and almoſt unſurmountable ones. This has been particularly the caſe with thoſe who have attempted to reſtore the Digamma to the poems of Homer ; a taſk certainly of extreme nicety and difficulty, but which will, I hope, be yet found praQticable ; for, until it is accompliſhed, the mi- nuter beauties of his poetry, ſuch as elegance, purity, and correctneſs, in which it excels as much as in ſublimity and expreſſion, muſt remain con- ecaled from the generality of his readers.

When the ſagacity and erudition of Dr. Bentley had diſcovered the want of this letter, Dawes, who, like many others, borrowed his ideas, and repaid him with abuſe, aſſumed the taſk of pointing out the words to which it ought to be added, and the figure by which it ought to be repre- ſented. In the latter he has been proved to be miſtaken, as Bentley has been proved to be right; but in the former his authority is ſtill held in high eſteem, though but little deſerving it; for he has raſhly foiſted in this aſpirate wherever the metre ſeemed to him to want propping, without examining whether or not its power was ſuch as the place required, or the etymology of the words admitted; whence he has brought this branch of criticiſm into ſome diſgrace among the learned in other parts of Europe;

(1) See Mr, Wood's Plan and Deſcription

4 who,


= \ 2 — 4 .

E i en eds bf 2 ee ENG

pronounced in two ſyllables,


who, with the natural prejudices of pedantry, have pronounced che en-

quiry to be vain, becauſe it has not been purſued with ſucceſs (1). The metrical power of the olic F is almoſt, if not is; 15

4 as that of the H or F; for it is equally a ſimple or vowel aſpirate,

pronounced with nearly the ſame degree of conſtrained expiration, and, in

the Zolian dialect, often occupied its place, or, at leaſt, the place which

it held in the Attic dialect; the tranſition. being extremely eaſy, in a lan- guage not fixed by any decided principles of orthography, from one letter

to another, when both are of the ſame claſs, and poſſeſſed nearly of the

ſame power (2). The Pelaſgic clans of Italy ſeem to have employed it occaſionally as a vowel, the antient medals of Capua being inſcribed IAA (3). It is poſſible, however, that the name of this city was then

or Zfolian Digamma, is invariably uſed as a vowel in that language, though employed as a conſonant by the Perſians, who have corrupted it

Vau or Waw. There is certainly no reaſon why the Campanians might not, in their dialect, have uſed the correſpondent letter as a pure vowel, though the other natives of Greece and Italy employed it as a pure aſpirate, ſome with the metrical power of a ſingle conſonant only, and others with

the general metrical powers of almoſt every other letter, as I ſhall now f proceed to ſhew, 5451 .

(1) See D'Orvill, ad Charit, p. 202; and Erneſt. ad II. n. 172. (2) See Salmaſ. in Crenii Muſ. Philolog. & Hiſtor. p. 78. In the Heracltan tables the ſame word is written with the F (in the Pelaſgian form L) when alone, and with the F

when compounded; as LETOE and NENTAFETHPIZ, which oceur invariably, The number

81x too, which in all other dialects is written & or LE, is in theſe tables EEE. rox is like» wiſe written in one inſtance Lor, and in two according to the uſual on: See I. 101, I22, and 127. Sod $4 x1

(3) Sce Comb. Pl, X1V. a

(4) Aſiatic Reſearches, Vol. I. Pe 30 & 31. i 4 . 2 F 2 That

KAPWA; but as the final A is never to be Found upon the coins, it is more-probable that the antient Oſc inhabitants did not employ it, but pronounced the name of their city KAP'HY, which a Welſhman would now write KArw. The Arabian Waw alſo, which has the ſame name, and probably the fame power, as the Pelaſgian Vau,

preciſely as we have the Roman V (4), which was originally the Pelaſgian

36 AN ANALYTICAL ESSAY That Mr. Dawes ſhould not have obſerved this univerſal power of the

  • FEolian Digamma is rather wonderful; for, beſides the analogy of ſound,

which might have led him to it, the very authority, which he quotes, aſ- ſerts and exemplifies it. The word conſonant, indeed, being improperly applied to this letter, might naturally have miſled a leſs acute and learned obſerver, but could not, one ſhould think, have miſled him, who employs much argument to prove it an aſpirate. The Latins,” ſays the Gram- marian Priſcian, in a paſſage cited by Dawes, ** employ the V for the Æo- % lian F, both being uſually Gmple conſonants, as in


© AT VENUS HAUD ANIMO NEQUICQUAM EXTERRITA MATE. The Zolians, however, ſometimes uſed the F for a double conſonant, « as in NEXTOPA AE FOr ITAIAOZ. In other inftances they uſed it 4 as a ſhort vowel, as in KAT XEIMA IITPTE AAFION ; and in others 4 jt has no metrical power whatever, as in AMMEE A' FFIPANAN TO

« AE TAP OETO MQEA AIFAIA.”” The reader, who is converſant with

the writings of Homer, will readily obſerve that this is preciſely the me- trical power of the other ſimple aſpirate, ſignified antiently by the figures H and g, and now by the mark (). Dawes would, indeed, ſubſtitute

the Digamma to this aſpirate in all inſtances where the vowel is lengthened

or ſuſtained by it; but as he has no authority for ſo bold an innovation except his own ſyſtem, which is contradicted alike by etymology, analogy, and antient monuments, his arguments, or rather conjectures, do not de- ſerve any very ſerious conſideration, He would even do away the autho- rity of his own quotation from Priſcian, by reading the paſſage of Alc- man (1), AMMI or AMME AE FEIPANAN, though the alteration, if other- wiſe admiſſible, would render the verſe totally inapplicable to the purpoſe for which it is cited,

To attempt to point out the inſtances in which the Digamma ought to be

inſerted in Homer, after the failure of ſo learned and ingenious a Critick,

muſt of courſe appear raſh and preſumptuous in one whoſe habits of life

(1) Aleman is ſaid to have been the firſt poet who employed any verſe but the Hexa- meter of Homer, Both his age and country are unknown, for, though he is generally ſaid to have been a Lacedzmonian, Velleius Paterculus poſitively aſſerts that their pretenſions to him were ill founded. See Meurſ. Miſcell. Laconic. Lib. IV. c. xvii.


88 nne n RE ES: erent RR

* ” . 4

D Ne 8 n

  • te by 6.7 od? . 4 22 B - * ante I mw „ 2 dey

0 *


have not enabled him to apply his mind to the ſubje& with the unremitted > diligence of a proſeſſed ſcholar. As, however, I may throw out ſome hints which may excite the curioſity, or guide the inveſtigations, of more learned perſons, I ſhall offer my conjectures in as few words as poſſible. To do this

with that method which is equally requiſite to conciſeneſs and perſpicuity, it will be neceſſary, in the firſt place, to take an accurate view of the

flexions of his words, and to conſider them as written in the characters

which he employed, or which were employed whilſt his language was the familiar vehicle of ſocial intercourſe among his countrymen, and had not been conſecrated by the ruſt of time to the ſole uſe of poets, who employed

it only upon his authority, and when writing in his own metre. Not that I would infer, that the ſtyle of Homer was what we ſhould now call

obſolete (that is, ſo obſcured by time as to be intelligible only to the learned) at any period of Græcian literature; but that many of his words and flexions, having ceaſed to be in familiar -uſe before even the commence- ment of proſe-writing, were ever after reſtricted to the Heroic or Hexa- meter verſe, and not allowed even to the ſublimeſt dramatic poets who em- ployed a different metre, though the loweſt of the audience would have found no difficulty in underſtanding them. The true meaning and ety- mology, indeed, of ſome of his words, was loſt ; but cuſtom had ſupplied another which every one knew.

In the variety of the antient flexions conſiſt the dialects of Homer, which

muſt not be underſtood to have been, in his age and country, provincial»

iſms, like the dialects of modern Italy, but merely variations upon one tongue, all equally authoriſed by general uſe. Some of them, indeed, might have become provincialiſms, even before his time, in particular parts of Greece; but, nevertheleſs, the mixture of the olie and Ionic emigrants muſt have again confounded them in Aſia, and rendered them of general popular uſe before he wrote; for we may conclude that, as his

poems were addreſſed to the general maſs of mankind, and are remarkable,

above any thing, for extreme perſpicuity, his words and n. were all ſuch as every hearer would readily underſtand.

Theſe antient variations or dialects conſiſted chiefly of different modes and degrees of aſpiration, and the broad and ſlender enunciation of tone

ſignified by the vowels A and E; which, though originally differences of

3 irregular




irregular licence, were, by degrees, as the language became ſettled, trans formed, by accidental habit or faſhion, into particular provincialiſms, diſ- tinguiſhing the olian and Ionian Greeks, whoſe ee the pa- rents of all the reſt; for the Doric is principally a con ion of the Folic, and the Attic of the Ionic (1). Homer, I am inclined to think, was equally unacquainted with both theſe provincial contractions; for, though Atticiſms occur very frequently in his works, as we now have . them, they appear to have come from the Athenian 'and Alexandrine edi- tors, through whoſe hands they paſſed in their way to us. Not but that contractions and eliſions were in uſe even in the earlieſt times, but they were entirely different from thoſe which characteriſed the Attic dialect. From the ſame corrupt channels flowed the anomalies and poetical licences which commentators have pointed out and explained, but which were cer- tainly unknown to the pure and regular diction of the poet, as will more fully appear from a ſhort analyſis of his flexions, which are all upon one principle, though claſſed and ſubdivided by grammarians and ſchoolmaſters, for no other apparent purpoſe than to load the memories, _ perplex the underſtandings, of their pupils.

It has been obſerved by Dawes, that the nouns ending in -E antiently ended in -EFZ, from which their oblique caſes are regularly formed, as FINNEFZL, -Erox, EI, -EFA, -EFE (by eliſion EF), -EFE, -EFOIN, -EFEZ, -EFQN, -EFEZI (contracted to -EFEI), -EFAE ; each of which ſuffered various contractions in later times; but in Homer the loſs of the Digamma is almoſt conſtantly ſupplied by the Epſilon being tranſpoſed into an Eta; nor do I know of more than two inſtances in his works of an oblique caſe, or plural number, remaining without the aug- mentation of an additional ſyllable, Theſe two are the words ine (2)

(2) Though theſe four are the only dialects that were regularly cultivated and fixt, many more exiſted in the licentious variations of ſpeech that took place through the wide diſper- ſion of the Greek colonies. Herodotus mentions four different kinds of Ionic ſpoken in

. Afia only *; and it is probable that, before the Macedonian conqueſt, almoſt every ſtate had ſome peculiarities of its own.

(2) Ixus & lng ire 3. eh wero en. II. A. 151. 

  • Lib. I, 8. 1462.

1 l 3

— 1 y YN Br 7 ” 2 r= the 3 11

45 F


and ode (1); to which we may add from Heſiod, whoſe poem equally, requires the Digamma, a third, Barn Theſe are all contractions f the antient forms; but in what manner they were antiently written is difficult to ſay ; for, though both the T and the I were employed to replace the F, we cannot reſtore this letter without reducing the nominative and genitive ſingular, and the nominative plural, to the ſame form, only diſ-

_ eriminated by the circumflex. There is, however, no doubt but that, in

the nicety of antient pronunciation, this circumflexed form was diſtin- guiſhable by the ear as well as the eye; wherefore I am perſuaded, that the primitive contraction was from -EFOE and -EFEE to EFZ, changed in the genitive to -, and in the nominative to -«5, and afterwards, by the Attics, to -ys ; by which means any ambiguities which __ have ariſen were avoided.

Upon the ſame plan the oatranymice, and other words of ths fline claſh; ending in -A, -AE, -EZ, or -HE, ſeem to have been declined, except that the aſpirate was uſyally dropt in the Ionic pronunciation; whence, when the penultimate ſyllable is long in the oblique caſes, they are always in the ZEolic, and, when ſhort, always in the Ionic; for the /Eolians retained the uſe of the Digamma after it had been neglected by the other Greeks, whence it was called Æolic. Ihe genitives Arge3ao and Inna abſolutely xequire the inſertion of the aſpirate, in order to give the penultimate vowel its due length, and were, therefore, undoubtedly written ATPEFIAAFO and IIHAEFIAAAFO; but Arpt and xe require its omiſſion, others wiſe the two laſt vowels could not coaleſce into one ſyllable as they uſually do; wherefore they muſt have been written ATPEFIAEO and ITHAEFI- ako, the two firſt Digammas in the middle of the words ſtill remaining, as they belong to the roots ATPEFZ and IIHAEFE. In words of the for- mer claſs too the Digamma was retained even in the flexions through both

(1) Odvoiv; . Na xugs eig in xagty. od. N. 397» We have, indeed, the accuſative vv in another inſlance, which is generally ſuppoſed to be produced by an apocope of the laſt letter, u' avr' ay yin em: Tvdn gear Axon e (II. 0 384); but the ſingularity of this form renders it ſuſpicious, and a ſlight alteration in the order of the words makes it regular EN av TYFAEF' Eu! ANTEAIHN EZTELAAN AxAtol. Every one, who has examined the various readings, knows how often words have been tranſpoſed to the detriment of the en and metre in the manuſcripts and old editions.


- -


dialects; whence we have NAP and NEF, or, as they are now written, yea and vv, with their correſpondent forms in the oblique caſes always in two ſyllables, whereas the nominatives are always in one, which proves that the aſpirate was retained in declenſion. The firſt ſyllable is, indeed,

ſometimes long and ſometimes ſhort, the ſhort vowel before the aſpirate

being pronounced either way; whence we have »y« and yea, both of which ought to be written NEFA or NAFTA. The Ionians did, indeed, in ſome

inſtances, drop the aſpirates, and extend the vowels, contrary. to etymo-

logy ; but it is very uncertain whether this provincial innovation prevailed at all ſo early as the time of Homer, and very improbable that it ever pre- vailed in the declenſions of the nouns,

Whether the Attic or Tonic terminations of the patronymics, &c. in -Hx, and the formation of the genitives in -«@ or -EO was at all known to

Homer, I have ſome doubt, as the ZEolic terminations in-Ax and -AFO

ſavour more of antiquity, and the latter might have been reduced to one

ſyllable, AF, by the eliſion of the laſt vowel, which we know was prac- tiſed, even before the Digamma became obſolete, to form the Doric geni- tive in A, which occurs on the very antient medals of Thebes and Mace-

don in the names EYFAPA, AMYNTA, IIEPAIKKA, &c.(1) In the ge-

nitives plural of the ſame claſs, the Attics omitted one vowel, and the Do- rians the other; whence we have EIKEAIQTAN, ITAAIQTAN, &c. in the one dialect, and EIKEAIQNTAN, ITAAIQNTAN, &c. in the other, both being contractions of the primitive /Eolian forms, EIKEAIQNTAFQN, ITAAIQTAFON, &c. originally, perhaps, written with the ſingle vowels EIKEAIAOTAFON, ITAAIAOTAFON, &c.

The aſpirates, both vowel and conſonant, were often elided even by the very antient Greeks ; whence we find ovg & be, e & y, &c. and the future tenſes of the verbs, ſometimes written with, and ſometimes without, the characteriſtic ZE, the omiſſion of which has cauſed the antient ſcholiaſts to miſtake them for preſent tenſes, and to ſuppoſe a ſort of licentious enallage, which, if admitted, muſt ſubvert all the principles of language. The Di- gamma was occaſionally elided in the ſame manner; but whether aurap, Gupu, muAvg, Gurus &c. were ever written AFTAP, AOFPY, IIOFATZE,

(1) See Dutens, p. 158; and Frælich. c. VII. :



later Attic for the F, it was alſo inſerted where that aſpitate never could

have been, and I believe, in ſome inſtances, reſtored t words from which

it had been dropt; for the diphthongs were much leſs uſed in the ſecond than the firſt ſtage of Greek orthography, whence we have EIMI in the firſt Sigean inſcription, and EMT in the ſecond; The negative O was at one period very generally written O; but it does not follow that it was ever written Or. In one inſtance Buacods occurs with the. firſt ſyllable

ſhort, whence ſome Manuſcripts have Boxeobe and GD gba.

E. 8'v u © 60s peubog eÞoySave, GANG Buneobes 4525 Aurov vd Cut, x. xiv rar mana - Ol. H. 387.

But though this elifion of the Y removes the metrical irregularity, the greater difficulty ſtill remains, for the word , as Clarke has obſerved, is totally incompatible with the ſenſe, which requires a conjunAive inſtead of a digunftive. I would therefore read, SEES

Es * d uur ods Aobo ada dar, 7" Kats rer

BuAzobs gu U £810 WETpWG, WAYT Ge

Though the Digamma, as well as the other aſpirates, — be. thus

elided, no licence could ever add or inſert either into words to which they

did not regularly belong. The antient ſcholiaſts and grammarians, indeed,

who wrote ſo many ages after the two vowel aſpirates had both been dropt

from the Alphabet, and the one: wholly obliterated and diſuſed, finding that, which was retained in pronunciation, ſignified, when ſignified at all (1), only by the inverted comma (), confounded it with the accentual marks, and eſtabliſhed certain whimſical rules of their own for affixing or omitting it. hey #) The Alpha before a Delta, they decided, ought always to be acpinmedh, unleſs it was a craſis, or ſignificant of privation (2); but no vowel could be aſpirated in any caſe if followed by an n conſonant and a P,

(1) See Euſtath. in Odyſſ. z. VC. 161, where it appears that the majraſaipls which he

uſed, though he wrote as late as the twelfth OT had no notes of aſpiration. See alſo Erneſti ad Loc.

(2) Schol. Ven. ad II. A. 88. BY Vole

G _ whence

  • _

F * „ 1 "na. 4. =

  • .

nan bee. as Dawes and bie learned and ingenious editar have ſup- poſed, I much doubt; for though the T'was very, generally inſerted in the


whence gp, ape, ag &c. ate without it (1). The A alſo could neyer be aſpirated if followed by a A and a dental or palatial conſonant, whence ro is formed from dau (2) ; followed by a P and M it is, ene to be aſpirated, though there are ſome exceptions (3).

The a: diphthong, beginning words of more than one ſyllable, was ne- ver to be aſpirated (4); and the Z, followed by an aſpirated conſonant, was ſuppoſed to prevent a vowel preceeding from being aſpirated, whence the T in «obog is lender, though in the verb from which it is derived it is aſpirated (5). In mdprrrog it was alſo to be flender, though aſpirated in eg, becauſe followed by a A in a word the third ſyllable of which was a pure vowel (6).

Some Criticks were for aſpirating the augment 1 in 8 tenſes of

particular verbs, and others of others (7). Ptolemy of Aſcalon decided that the inſertion of the I ſunk the afpirate;

whence dg became Ng, and adde, when joined to the adverb tu, wal (8).

According to Dionyſius of Halicarnaſſus, the aſpirates coming in the middle of compound epithets ought to be preſerved, but clided. in proper names, whilſt Herodian maintained the contrary (9):

Some, however, of the more antient grammarians underſtood the prin- ciples of their language better; and it appears from ſeveral paſſages of the Venetian Scholia, that Ariſtarchus and his followers were for aſpirating all words according to their etymology (10). That this is the true opinion, we might venture to decide, even if it was not ſupported by ſuch reſpect- able authority; for, as the ſimple aſpirates were originally parts of the Al. phabet as much as any other letters, it is natural to ſuppoſe that they were employed upon the ſame principles as the reſt. They were, indeed, more.

(1) Schol. Ven. ad II. M. 391.

_ (2). Euſtath. p. 146, & 766; J. 41+ (3) Ibid. p. 140, 11. (4) Ibid. p. 1626, 1. 38. (5) Ibid, p-. 1431, 6. (6) Ibid. p. 345, 1.

(5) Ven. II. M. 55.

(8) Ibid. x. 340.

(9) Ibid. o. 750. (10) Ib. O. 365; and a. 235, and 247


flexible than the conſonants or liquids, and therefore more liable to local and habitual variation and corruption; but, nevertheleſs, leſs ſo than the vowels, which were conſequently more varied than either by change of dis. alect. The Heraclean tables, which apppear to have been written juſt when they were falling into diſuſe, are more licentious and irregular in the omiſſion and inſertion of them than any other antient monuments extant. One inſtance has been already given, and we find another in the verb xu, which is ſometimes, both when alone and when compounded, written FEXQ(1), and ſometimes EX (2). Many other words alſo, which are every where elſe unaſpirated, are here uniformly aſpirated, ſuen as OK To, FENNEA, FAK POE, FOIEQ, and FAPNHZIEZ. Others, on the contrary, which are every where elſe aſpirated, are here unaſpirated, ſuch as OPOZ, AAIA, AMAZITOZ, and AEKOMAI, The cuſtom of continuing the aſpirate at the beginning of a word, when it has been added to the pre- ceeding letter, appears from theſe tables to be modern, as we find x' TO, and not X FYIO, thoogh this prepoſition is in every other inſtance aſpirated. 1 According to the antient wisse of declination, the Digamma appears: to have been the characteriſtie letter of the oblique cNes in the maſculine and neuter words terminating in -O and -Z, and the feminine in -Q, -A, or -YE, and -A or -H, though it is only wanting to ſuſtain the me- trical quantity in the ZEolic genitives plural of the laſt, as MOFZAFTN. NTMoAF aN, &c. The general analogy of the lauguage, however, makes it probable that it originally prevailed alike through all, and that -OFO, was the /Evlic termination of the genitive ſingular of maſculine words in

Or, as -OIO was the Ionic; which, being both gradually changed by

the contractions and eliſions common in the Greek language, became OO, -O, and - Or, the laſt of which was probably firſt written OF, for no re- gular proceſs of etymology could have placed the I here; though, as this vowel was very generally ſubſtituted for the F, when it fell into diſuſe, we may reaſonably ſuppoſe that it was ſo. in the preſent inſtance, Even in the modern nd of Hon. the Lad of the NN name Ti]eog

(1) Tab. Neap. I. 59, 725 82,

(2) Ibid. I. 43, 68, 69, 73, 93, 109. ;

G 2  a is 


their notions of arbitrary extenſions, pleonaſms, adſcititious vowels, &c.


is Heriwe, which, I think, can only be a corruption of METEOFO ; and not, as the Scholiaſt explains it, an Attic extenſion of the penultimate, and pleonaſm of the ultimate, vowel (1) ; ſuch arbitrary extenſions and pleonaſms being, I believe, wholly unknown to the Poet; whoſe words, though frequently .contrafted, were never amplified or prolonged but ac- cording to the ſtrifteſt rules of etymology. This, I believe, may be, with equal truth, obſerved of the words employed by all corre& writers in all languages; for, though ſome degree of licence in contracting and abbrevi- ating is allowed in all, there is none that I know of which admits of any licence whatever in extending or amplifying. I and ae, each of which occur only once with the penultimate ſyllable long, though ſo often uſed with it ſhort, are probably remains of the ſame antient flexions ; for the laſt ſyllable in both is long by poſition, 1A wporapoiber and ava x ra- loo, which I would write FIAIOFO IIPOITAPOIGEN and ANENZIOFO' KTAMENOIO,

The Cratylus of Plato ſeems to have furniſhed the antient ſcholiaſts with

Ke. What the Philoſopher meant by that dialogue it is difficult to gueſs, for there is no appearance of humour or irony, and yet the etymologies' which it contains are infinitely too abſurd for any man of common - ſenſe ſeriouſly to have believed. Every cobler at Athens muſt have known that- ande was not derived from arri and pro, nor apabic from apa and Dog.) The reader who ſeeks for plain ſenſe, and not merely for fine periods, « cane

not but ſuſpect that Plato ſometimes wrote dreaming.

It is poſſible that the formation of the -O diphthong in the genitive caſe might have been, by corruption, habitually introduced, as well a8 by the regular apocopò or elifion of the ending vowel ; for we find the geni- tives in -EOE or -EFOL contracted to -OYE in the later Attic, or common Hellenic dialect, which can be accounted for by no rule or principle what- ever, unleſs we admit the metatheſis, or abitrary tranſpoſition ' of letters, which will. be conſidered in the proper place. The antient contraction was -OL; whence in the Sigèan Inſcription, which is Tonic, we have HEPMOKPATOZ for HEPMOKPATEFOZ ; and in the Sandwich, which is

(1) Schol, Ven, ad 11-4. 37a. at 1s) 3 Attic,


Attic, ENITENOE for ENIDENEOE. A ſimilar contraction occurs in the verbs, the ſecond perſon fingular of the imperfe& paſſive being changed from EO to -EO and -OY. In the Venetian Manuſcript we have Jagoiug, that is @APEEFE, inſtead of Yapovs, as in the common editions (1); which ſeems to be a ſpecimen of the primitive contraction of the regulas genitive APTEFOx, preſerved by accident, as the contraction OAYEEFE for OAYLEFOL, beforementioned, was by the metre. A corruption of the ſame kind as that which appears in the common forms of theſe words has taken place in the nominatives plural of the comparatives in -QN, as: APEIQN, the regular plural of which was APEIONEZ, contracted by the Tonians to APEIOEZ, pronounced in three ſyllables, and thence corrupted” by the Attics to APEIOYE. In Herodotus the contraction is more ſimple, and the comparative formed with the T inſtead of. the ; for the F; whence we find weve; for NAEFONEL or INIAEIONEZ, written by the Ro IMAEIOTE.

Theſe abbreviations have cauſed the poſitive and comparative to be a.  

times confounded in the flexions, as in XEIPEFZ, properly a labourer, or handicraftiman, but uſed figuratively to ſignify any private or common per- fon, the comparative of which, XEPEFQN or XE PEIMN, ſignifying com- moner, or worſe in general, and being contracted like other adjectives of the fame claſs, the regular flexions of the poſitive, ſuch as XEPEFI, XEPEFA,. XEPEFEZ, &c. now written. yepy, ien, X#gness &c. became miſtaken for abbreviations of it, and, I believe, ſtill continue to- be ſo, though the- ſenſe of the context will eaſily point out the difference. The word XE- PEFSZ having grown obſolete at a very early period, whilſt its comparative- continued in general uſe, very naturally cauſed the confuſion,

Hoc, the genitive fingular of eus, is a dactyle in Od. Z. 303, notwith-- Sanding: the double vowel in the penultimate; which proves that it was an-- tiently written FHPOFOFZ, the penultimate of which might, be pronounced

either long or ſhort. It was probably from not underſtanding this general

principle of the antient flexions, that the rath grammarian Zenodotus. would have changed yopyw, Yep, to yopyur, yoeyorcs (2); for, had he un-

(3) II. r. 573. (2) See Schol. Ven, in II. e. 349




derſtood this part of the analogy of his own language, he would have per- .ceived that yopyu; was the regular contraction of the regular genitive Op- rorox, except that an obſolete letter was changed for a common one. Mo dern interpreters ſeem to have erred in the ſame manner when they con- found a, that is ALAOFA, the accuſative ſingular of AIAQE, reverence,

or viriuous ſhame, with a contraction of AIAOIA, the private parts, by which means they render obſcene and ludicrous one of the moſt pathetic and ſolemn paſſages of the Iliad (1). This error ſeems to have originated from the blunder of a tranſcriber, who, in another paſſage, has put u. for ad, and has been followed by all the editors (2). 1

The accuſative plural of the maſculine words in -O, and feminine in Tx, ſeems to have been formed by a change and contraction fimilar to what has taken place in the genitives ſingular and nominative plural above- mentioned; for Moyus ſeems equally to ſtand for AOFOFE; ed for EPINNTFE, the contraction of EPINNTFAZ ; and ur, for KAITTFE, the contraction of KAITTFAE ; though I believe this laſt word ought to be written at length in every inſtance where it occurs, and the firſt ſyllable pronounced ſhort, as it is in Euripides; and alſo in Homer in other words derived from the ſame root, ſuch as KAIEIH, KAIZION, &c. We have, indeed, the accuſative u in O8, E. 470; but this ſeems to be equally a contraction of the antient accuſative KAITTFON, which prevailed through all words of this claſs; whence the vowel is now ſuſtained before my . ogg ru ca] / once probably written FO ITYFON KAMAEHI. The laſt ſyllable of the coutracted form of the accuſative is always long, be- cauſe, in antient orthography, it was TFN inſtead of -v. In II. 6. 318, the penultimate in a genitive ſingular of a word of this claſs is long -

e' Ac hich might have been antiently written and pronounced KEILES' n' FIATFOE, or KEIZEe TIIO FIATFOE ; for though the firſt ſyllable of FIATE is uſually long, there is no reaſon from analogy why it ſhould be neceſſarily ſo.

The earlieſt inſtances which I Wea met with of genitives in -O are upon the medals of Dionyſus, King of Syracuſe, and Alexander II. and

(2) IL X. VI. 35. (2) II. B. 262. 7

— Philip,


Philip, the ſon of Amyntas, Kings of Macedon. I have, indeed, ſeen in

books EYPAKOZIOT. FTEAQNOE upon the medals of Gelo, King of Syra»

cuſe, who flouriſhed near an hundred years before any of the abovemeu-

tioned princes; but upon inſpecting the original coins, of which I have ſeen vaſt numbers in the different cabinets of England, France, Sicily, and

Holland, I have uniformly found EYPAKOEIOI FEAQNOE. This has

given me ſome ſuſpicion of the medals of Dionyſius, of which I have not ſeen any with the name at length; but nevertheleſs, as both the kings of Syracuſe, who were ſo called, made the Attic dialect the language of their court as well as the kings of Macedon. abovementioned, it is poſſible that

they equally employed the termination in -OT, which peculiarly belonged to it, but which does not appear to have been employed even at Athens till afterwards ;, for. it is not to be found in the Sandwich inſeription, . Which is public act of a later date. Probably the orthography of the Attic diale& was firſt adapted ſtrictly to its pronunciation in theſe courts, where, as we are informed, the moſt powerful ſovereigns of Europe thought it an object of ambition to be able to ſpeak and write it correctly. This may account for its being formed with ſo little attention. to etymology.

This diale& was, at that. period (about: four hundred years before the

Chriſtian æra), becoming every where the faſhionable language of letters and philoſophy, owing to the well-earned reputation of the Athenian wri- - ters, which having ſoon after recommended it to the patronage of the great conqueror of Perſia and his ſucceſſors, it became the general language of

civilized men, and was thence conſidered as. the. common Hellenic dialect,

and the ſtandard for purity, though it is in reality one of the moſt corrupt

dialects, as far as corruption conſiſts in deviation from primitive roots.

Whether the word EXPYKOEIOL upon the coins of Gelo be a nomina- 

tive plural, or an abbreviation of the genitive ſingular, is difficult to decide,

though I think the latter moſt probable. There is not indeed any inſtance

of ſuch a genitive; but nevertheleſs, by the ſame rule of analogy that OO is abbreviated by the apocopè to -OF, - Olo may be abbreviated to.

Ol. The moſt faſhionable and poliſhed diale& too, in the time of Gelo, was that of the Aſiatic Ionians, which employed the I rather than the T inſtead of the F; for the I was their uſual ſubſidiary letter, as appears, not only in the r . ſuch as A0 O10 for A0 OFO, and ZEID -

4 A=

er bet.





'EEIO for ZEFO, but alſo in variations of a more ſtable and permanent kind, ſuch as PEIA for pea or PEFA, KPEIQN for xpiwwy or KPEFQAN, &c.

It was alſo employed for the F and T by the Dorians; and, on ſome oc-

cafions, by the /Eolians, if the preſent orthography of the fragments of Sappho, &c. is to be relied upon, which I cannot anſwer for. In the flexions it was inſerted or omitted arbitrarily, even long after the dialects

had become eſtabliſhed provincialiſms ; whence we find upon all the ſilver

medals of Agathocles, of which great numbers are extant, the genitive

caſe of his name written ATASOKAEIOE, whilſt the gold and braſs, the latter of which are equally common, have uniformly ATA®OKAEOE; ſo

ſo that the accidental or habitual practice of different mints diverſified the orthography even in the ſame country, and under the ſame prince (1). The antient folic termination of the genitive of nouns in -OE ſeems to be preſerved with but little variation in the relative pronoun, even in the preſent corrupt ſtate of Homer's poems; for I think 5 cannot be de-

rived from & or FOE any otherwiſe than by being a corruption of FOFO,

whence the laſt ſyllable is never long but when rendered ſo by poſition (2). The poſitive pronouns EY and FO were alſo declined upon the ſame plan, as appears from the genitives ev and zu, evidently corrupt abbrevia-

tions of TEFO and FHEFO, often written at length, in the Ionic manner,

ov and % The accuſative is ſeems likewiſe to be the Ionic mode of

writing and pronouncing the antient regular accuſative FEFA. The nomi-

native plural and dative ſingular and 6 belong to another declenſion, and are only diſtinguiſhed from the correſponding caſes of the relative pronoun by the accentual marks, which were not invented till the end of the third

century before the Chriſtian æra, and not in general uſe till the middle

ages (3). I ſuſpect, however, that this dative fingular has been ſometimes introduced where the old regular form EFI ſhould be; whence it ſome- times continues long before a word beginning with a vowel. The ſame may be ſaid of the dative «o,, of which the laſt ſyllable is ſometimes long in the ſame predicament, and ſhould then probably be written EMEFI. or

(1) The 1 was very generally added to the E by the early Greek writers, as the T was to the O. Euſtath. P- 511, I, 1.

(2) o «Avg, Il, B. 335, o greg, Odyſſ. A. 70. (3) See Villoiſon. Prolegom. in Homer, p. 12.

4 EMEF*,


written ehe and epeiv.

In the dual and plural anmbers the o has taken the place of the p, and

the hiſſing dental aſpirate been prefixed to the third as well as to the ſe- _ cond perſon, which it probably was origiually iu the ſingular, atleaſt in

ſome dialects, for the aſpirates were changed even from the T to the F.,

that is, from the harſheſt to the ſofteſt, by the variations of diale& (1). Hence we have owe or c, 'oÞww Or cn, oÞ5i;, cds, eie Or ch, and opa;, Which ſeem to be only corrupted: contractions of EEFOE, EEFOIN, EEFEZ, ZEFQN, EEFIEI, and EEFAZ, though it is probable that they had been adopted by general uſe even before the time of Homer.

  • Dawes would prefix the F both to the relative and poſitive pronouns, in

contradiftion to many very antient inſcriptions, and without any ſupport from analogy, merely becauſe he thought the F inſufficient to ſuſtain the metre; but the very authority which he cites proves that the metrical power of both the ſimple aſpirates was the ſame, and that his conjecture

was therefore founded upon a falſe ſuppoſition. In a verſe, indeed, of the

ZEolian poet Aleman, cited by Priſcian in the paſſage before quoted, the pronoun poſſeſſive begins with the Digamma (FOr); but in Homer the ſame genitive is du, 8040, and ines, Occaſionally contracted to ; whence it ſeems that the word was FEFO Z, the regular adjective of O, which was declined ſometimes like the neutral, and ſometimes like the maſculine, nouns, in -OZ, -FEFEFOZ, or FEFOFO and FEFOIO contracted to PE- FOF, written in modern orthography - zs, and, by an elifion of the firſt ſyllable, common in the folic and Doric dialects, FOF, which, by a

change of the aſpirates, became & or FOY. The declenſion after the man-

ner of the neutral nouns in -O is rejected by the authors of the Venetian Scholia, who explain ines to be the genitive fingular of tvg, good, the geni- tive plural of which, pronounced after the Æolie manner, often occurs, tawy or FEAFAQAN. Others of the antient editors wrote 30 inſtead of i

for which, however, it does. not appear that they had any authority. Pro-

bably the opinion of the ſcholiaſts is right, and in that caſe «ug and e a” always to be aſpirated, FEYE and FEY, which may account for the

(1) See en magn. in W yo 3 and Villoiſon. 1 1 Homer, p. 2.


EMEF', conſiſtently with the antient ne, -EMEFO and EMEF, now =

H TE firſt

=D A


firſt ſyllable in the diæreſis being ſometimes long and ſometimes ſhort, and alſo for the concluding vowel of the preceeding word being frequently ſuſ- tained, It may alſo ſhow us the true meaning and etymology of the Latin appellative uxus l which ſeems exactly to correſpond with the zu BIEN! of the French. FOFOE is formed from the relative pronoun as FEFOL is from the poſitive, and declined and contracted in the ſame manner,

I have often been inclined to ſuppoſe the paragagic particle 9, a corrup- tion of the antient dative caſe, and to think that G., pero, oyeopr, &c. were once written BIEFI, ETPATOFI, OXEFLI for OXEFEEI, &c.; for, beſides the inſtance of the pronoun abovementioned, we find how eaſily the F became a © from the preſent practice of the modern Greeks, who terminate the words, antiently ending in EFZ, and then in -ETZ, in c, as BATIAE FZ, afterwards BATIAE TE, and now Barhegg. If, how- ever, this was originally a corruption, it muſt have been authoriſed by ge- neral uſe even before the time of Homer; for in his works the termina- tions in e are employed in a manner adverbially to ſignify both the geni- tive and dative caſes,

The dative plural of neutral words in - OS has frequently the penulti- mate of the antient form ſhort, as ub for TTHSEFZI, or, as it is now written, gy0eoo: ; in which caſe the aſpirate was elided, as even the leſs pliable conſonants frequently were in inflexion; whence we have the ab- breviated comparatives beforementioned, and alfo the oblique caſes of other words, formed upon the ſame plan, ſuch as xuz«w, that is KYKEIOA for the regular accuſative KTKEIONA; x or IXOA (for ſo it ought to be read according to the Venetian Manuſcript and Scholia) for IXOPA, the

accuſative of IXQP (1). Where, however, the F is in the nominative, it

is rarely, if ever, elided in the oblique caſes, wherefore, inſtead of wyaiog vii, which occurs only once, I would venture to read NHAEFOL ETIE; for, though the firſt ſyllable of EYIOE is uſually long, it is not invariably ſo, and I think in this inſtance ought to be pronounced ſhort, as in II. E. 612, in order that the I might be added to the E; which may, neverthe- leſs, be rendered long by the ſucceeding liquid M. The genitive z7yA0g might alſo have been written NHAEFE, like OATLEFYL, before conſidered,

(1) Il, E. 416.


PP 4 -


Although an aſpirate may extend the ſucceeding as well as preceeding vowel, provided it be in the ſame ſyllable, it cannot, when placed between two ſingle vowels, render both long; for, if it be not dwelt upon, both

will be ſhort; and, if it be dwelt upon, that alone to which it is added

will be long. Hence we find, in the modern orthography ee and veg,

uporcoros and xporiwveg, &c, but never yyws or ug, becauſe in the original flexions, "NEFOE and KPONIFONOZ, the F might be added to the pre -

ceeding or ſucceeding yowel arbitrarily, but could not be added to both at

once. We have, indeed, Ilegoys and Ngiurog, which muſt have been equally written NEPZEFA and QPIFONOF, but the aſpirate and liquid,

preceeding the ſecond vowel in each, are ſufficient to extend them, ſo that the F may be added to the third,

This effect of the F ſeems to have continued after it had ceaſed to be in uſe; for, in the oblique caſes of this claſs, the preceeding vowel being long, according to the old Tonic pronunciation, always makes the ſucceed · ing one ſhort, even in later writers, and the ſucceeding one being long,

according to the Attic pronunciation, equally makes. the preceediug one

ſhort. The converſe, however, does not hold good, for either of them being ſhort does not neceſſarily make the other long. Hence we find in the ſame paſſage of the Odyſſey N and Nye, and in the Attic writers

uniformly NN, Oncia, Av, &c. which Homer never employs be- cauſe incompatible with his metre. The word aupnerpic, however, ſnews

that they were not inconſiſtent with the cuſtomary pronunciation of his age and country, as moſt of the Attic peculiarities were. Ihe vowels being thus arbitrarily extended by the aſpirates and liquids muſt be underſtood as a fundamental principle, but not as invariably ad- hered to in practice, for local or temporary habit had fixed the pronuncia- tion of particular words to one mode even in Homer's time. Thus the adjective KAAOE has the firſt ſyllable invariably long in the Iliad and Odyſſey, and invariably ſhort in the Attic writers; whilſt Heſiod, Theo- critus, and other later poets, who employed th dialects more un make it either long or ſhort, · as ſuited their purpoſes. The final. A of feminine words, ſuch as @EA, &c. .ſeems to have 5

rendered long merely by the emphaſis or cuſtomary pauſe uſed in ſpeaking,

for more is no authority, either from etymology or antient monuments,

H 2 „„ 7), oo OT 



which can juſtify the inſerting the aſpirate or doubling the vowel. The ſame may be ſaid of the terminations in -H, which in all very antient in- ſcriptions is -E, though it was certainly pronounced uniformly long.

A very learned and ingenious perſon has attributed the extenſion of the vowel before liquids and aſpirates to a ſimilar cauſe, that is, to the muſical pauſe or czſura (1), which muſt certainly have had greater influence upon the very antient verſe, that was always chanted to the ſound of an inſtru- ment, than upon that which was intended merely to be read, That this

pauſe did regulate the actual quantities of thoſe ſyllables, which were com- mon from their poſition, ſo far as to decide whether they ſhould be pro- nounced long or ſhort in each particular inſtance, I have no doubt; for, as the learned author has obſerved, they are never extended but when begin- ning the foot where the pauſe naturally took place : but that this pauſe could ever make a ſyllable, ſhort by poſition, long, I can ſcarcely admit; for the few inſtances which occur in Homer of the ſingle vowel A being pronounced long before the ſingle mute conſonant Il in the compounds of the prepoſition ANO, are not ſufficient to eſtabliſh a general concluſion, as

ſo trifling a licence might have been thought juſtifiable in works ſo long 

1 and ſo finiſhed ; or might even have been intentional irregularities, intro- duced to break the uniformity of the Hexameter verſe in the ſame manner as the gv axe@puxc, or verſes beginning with a ſhort ſyllable, As the nice ears of the Greeks abhorred the concurrence of conſonants, they altered many words, the original forms of which are, however, pre- ſerved in the oblique caſes, and in the Latin. The participles in -AL and EI ſeem to have once ended in -ANE and -ENL, like the Latin, whence the regular oblique caſes are in -ANTOZ, -ENTOL, &c. HA leems alſo to have been originally NANE, from which all the oblique caſes now in uſe in the maſculine and neuter genders are regularly formed, except the dative plural, which has become m4o:, though the primary form AN- TEZI or NANTEZEI is preſerved in. Homer, who, when he employed the contraction, probably employed the ſimpleſt and moſt direct, NANTEL, Dawes would, indeed, ſubſtitute the Digamma to the conſonants, and write NAFLEI from HAF; for which there is no authority but the ana-

. us 4. 7 + K R

4 1 » ' *

2 ?

(i) See Lib. ſing. de Rythm. Grzc. Ox. 789. logy .


logy of ſors words in which he ſuppoſes that aſpirate to have been inſerred to ſupply the place of elided conſonants, ſuch as odge or OaOFE for OAONY, and the terminations of the third perſons plural of the preſent. tenſe of the verbs, where the old Zolic termination ON TI, preſerved in the Doric, has been changed to -vo: or -OFLI. The words which origi- nally ended in - OF were declined like BOFE, BOFOE, or as now written Pres, Poog; and the oblique caſes in -OFE or - eu are contractions of- OFOE, as AFAQNE, -OFOL, OA contracted to -OFE or *vg and 3; RW -OFOE, -OFE, &c.

The participles of the preſent tenſe ending i in -QN, -OTE, or -OFZ, ſeem to have ended in -ONE, whence the flexion is the ſame ATAONTOE and Tru TON TORE, from ATAOTE and TTIITAN. The appearance of the T in the oblique caſes induced Dawes to imagine that it had originally exiſted between the N and the T in the nominative ; but iu this I believe he was miſtaken, for it is not authoriſed by the Latin of any period; and we find from the word AN AE or FANAKE, which formed antiently both FANAKTOE and FANAKOE (1), that the T was employed as a character- iſtic letter of theſe oblique caſes, as in thoſe of TAMA EQMATOYL, 'OTE OYATOYE, &c. It may indeed be ſaid, that the laſt word was originally. written OFATE or OFTZE; but, even if this be admitted, no ſyſtem- maker can transform TMA into MAT Z, for it belongs to a very nu- merous claſs, the laſt ſyllable of which is uniformly ſhort in all the antient poets, unleſs rendered long by the initials of the ſucceeding word.

The nouns in -IE ſeem to have been declined upon exactly the ſame. plan, and liable to the ſame variations, except that the A, inſtead of the T, was the characteriſtic letter of inflexion, as KAH FIE - IAO, the da- tive plural of which was contracted from KAHFIAETIL to kAHFIAZIT and K AHFIZI, or, as it is now written, u., in the ſame manner as AN-

TExI to ITANTET and HAI. Upon the ſame principles, EFTMEZI, or,

with the paragagic N, EYMEZIN, the regular dative plural of FTMEEx, or, as it was otherwiſe written, PTMEIZ, was contracted to EYMIN; or, as the Æolians pronounced it, without the aſpirate, and with a ſtronger

(1) Avaxo:y, Aiooxgpoive Schol. Ven. in II. a. 566; ſee alſo Euſtath. 142 5, 56; and He- ſych. The temple of Caſtor and Pollux at Athens was called the ANAKEION. : emphaſis



emphaſis upon the M, TMMIN or TMMI. The Ionian accuſative ſingulat ſeems to have been formed by a ſimilar contraction of an obſolete flexion, traces of which are preſerved in the Latin, where we find the genitives, which the Greeks terminated in -AOE, terminated in Dis, as PARIS 11s; and the accuſatives, which the Greeks terminated in -AA, termi-

nated in -DEM ; from which we may diſcover the old form in -A N, con- 

tracted by the uſual eliſion of the conſonants, and ſyncope of the vowels, to what is called the Ionic accuſative HAPIN, OGIN, &c. In the latter word, indeed, and ſome others, the contraction prevailed through all the caſes, whence ſome grammarians have made a ſeparate declenſion of them; but improperly, for, as Theodore Gaza has obſerved, all the names in IZ have their genitives regularly i in -IAOE,

It was probably from a view of theſe facts that Lennep, in his excellent little Book upon the Analogy of the Greek Tongue, has not noticed the conjecture of Dawes, but concluded that the terminations of the participles in -OYE and -QN were originally in ON, and that the ſubſidiary T and long -O were introduced merely to preſerve the due length of the ſyllable, when faſtidious refinement had dropt the conſonants. All the flexions of the feminine, and the dative plural of the maſculine and neuter, have been ſoftened upon the ſame principle; whence we have TYITOTEA inſtead of TTMNTONTEEA, TTHTONTEA, or TYIITONEA ; and TTOTOTEI in- Read of TYMTONTEEL, TYITONTE(L, or TYITONEI. We have alſo TT®OEIZA or FYO®OEELA for TYOOEN TEEA, TYOOENTEA, or TY» GENZA ; and TT@OEIEI for TY@QENTEZI, TYOOENTLI, or TTo- GENET, In the Doric diale&, the antient forms of the dative plural were preſerved, except that the E became an A, and the L was doubled, to ex- preſs the breadth and harſhneſs of this pronunciation. Hence, in the He- raclean tables we have IIOIONTATZIN, IIPAELONTAELI, FTHAPXON- TATTI, &c. which in ordinary Greek would be HOIOTZIN, IIPALEOT» II, 'YMAPXOTYEI, &c. That the F was ever employed for the T in theſe forms: is merely a ſuppoſition of Dawes, unſupported by authority or analogy, and probably untrue; for it is more natural to ſuppoſe that the Y was inſerted here, as in the inſtances beforementioned, by the reformers of the Attic orthography, who, when the quantities appeared defective through the eliſion of the conſonants, ſupplied them according to their 7 25 own



own pronunciation. In Homer the conſonants were probably retained, though the contraction had certainly taken place in the participles. In the third perſon plural of the preſent tenſe of the verbs, the termination, being ON Tx in the old olic and Doric, was probably -ONzI in the old Ionic, which being contracted to - OI or-, was again filled up, in the later Attic, with the Or diphthong, conformably to the pronunciation then moſt in faſhion, aud at length nn prevalent, tba: never juſtified

by ety Oe"

s EK G TI ON m.

HOUGH we cannot trace the antient orthography with the ſame preciſion by the mere rules of metrical harmony, as when aided by the regular analogy of the flexions, we have, nevertheleſs, in the extreme accuracy of the moſt antient poet, very plain directions to guide our en- uiries. 5 I, When we find a PRs vowel pronounced long, though followed by another vowel or ſingle mute conſonant, we may, unleſs in the inſtances already excepted, conclude that an aſpirate has been dropt, which we ſhall generally diſcover to be as requiſite to etymology as to metre.

Moſt of the following words have been remarked by grammarians for this defect, and, I believe, that the reſt were written vpon the ſame plan, and in the manner here propoſed: Apaiper—AMAIMAFQ, whence avoyuarjucicy or ANAMAIMAFEL and

fe, HY, probably written ANAMAIMAFETOE. AvaE———FANAKE, from FANAKQ or FANAEEQ, of which the imper- fect ſhould be EFANAZEE, and not were; which, as Dr. Bentley obſerved, never begins a line, becauſe the two. firſt ſyl- lables


lables in the time of the poet were ſhort. The Italian Greeks, according

to Heſychius, wrote it ANN AZ, in the Laconian idiom; and Homer has

the vocative avs or FANA from FANAE,

Amy—FANHP according to Dionyſius of Halicariaſfus (4 il); bot it does

not appear to have been ſo in Homer, for I do not recollect its

being preceeded by an open vowel unleſs where the paſſage is cor- rupt, as xa av3pur, which ſhould be KAEEF' or KAEE' AN APN, the antient accuſative plural of KAEOE being regularly KAEEFA, or KAEEA.

Aria, ai, avi ANIFA, ANIFQ, ANIFAAE(Q. The penultimate

being uniformly long proves it to have been written with the aſpirate, or diphthong ; and the moſt pro- bable etymology, given by Damm, favours the former.

Agy——FAETYT, being almoſt always preceeded by an open vowel.

Aare, ary——AFATQ, AFATH. In Pindar, as now written, auara, but properly AF ATA. In the, genuine parts of Homer it ap- pears to have been a triſyllable, as it muſt be according to

its etymology. The three lines alluding to the Judgement of Paris are

evidently ſpurious, being in every reſpect unworthy of the poet (2); and the other inſtance, where it is required to be read as a diſlyllable, Mr.

Dawes ſays, is to be corrected from the various readings; which I have

not, however, been able to diſcover, the line being in all editions the ſame :

— 'Os T4 H u or yopy Pero tppCanov arypiov a (3).

Perhaps for aypioy we ſhould read way, unleſs indeed AFATH might have

been occaſionally contracted to AFTH. In the ſame lliad (4) we have ao- Faro, or, as in other editions, aar, and aacaro, the metre requiring that the word ſhould form a daftyle— xa; yap In w wore Z aooaro, Toy wp apigoy, Clarke ſaw that this was corrupt, and therefore propoſed to read——xa; tyap oy wore Z auroro Tov wep apiger ; but, beſides omitting the particle u, which gives peculiar force and elegance to the ſentence, the.

in wore muſt neceſſarily be long before Z. I would, therefore, read

xc vg by vv wore Zu- Wome” oy Wep apigovz or, in antient orthography,

in 16, Ed. Hud, (a) 11. a. 28. ([) II. T. 88. (4) Vi. 25.

; KAI 


KAI TAP AH NT NOTE AZHN" EAFALZAO' ON nE APIETON, which gives both the ſenſe and metre correct and entire. Ariſtarchus is ſaid, by the Venetian Schotiaſt, to as read Tous for Zn" ; but I can ſcarcely cre- dit it,

He would alſo have ene the eee lite, but without "Y cient reaſons. His judgement, indeed, however good in regulating the minuter delicacies of compoſition, does not ſeems to have been adapted to decide upon the general ſenſe of a poet of ſo much ſentiment as Homer, otherwiſe he would never have thought of rejecting the four lines from the IXth Iliad, in which Phoenix mentions the deſign he had once entertained, in a fit of rage and deſpair, of killing his own father; for, without theſe lines, we do not perceive tl: intent of Phoenix's narration, which was, to ſhew the dreadful effects of anger; nor diſcover the cauſe why his father's houſe became hateful to him, which was, that it perpetually brought to his feeling mind the hideous ideas with which momentary paſſion had filled it. The cauſe of this ſtrange rejection was probably their having been puſhed antiently from their place by a really ſpurious line—Zewg rs xaraxlonug, xe £7 Lege which ſeems, by a ſpecies of advancement not un- common, to have ſlipt from the margin into the text, and by that means to have removed the pronoun fo far from the ſubſtantive, that, to preſerve its relationſhip, it was transferred over to the next repetition of it, The lines, as they now ſtand in Herglerus s edition, the only one that has re- admitted them, are,

— Harig à* dug aur ber

ox Ax x vνινανν , gUYBRY * emexexcner” pi

My Wore yuvaoiy 00w peroecl cc Praoy duo

EZ epebey yerywwra* Y 3” tre\uoy emropogy

Z.&vg re xarTax loving, Kai emaivy WEpTifoveirs

Erb“ ee £x874 Wap may e H & e DE?

Ilargos xw0jaevoro rare furyapa gp pan ha.. Tov fe 8yw “iM xatarkrapey oft xnrw Ana rig ab D wavy xonovs &; g wi Juprs Anus Ine pr, ras oveidics worn abr "Ng un Tarpeporog puter” ARα,zZ,ͤ x.



Inſtead of which they ſhould be, -

— ary * h ur olebiig Hex Aa xarypatro, guyipas d' drt Sh. My wore yu⁰νm in 01010 eo o800 ac Pinoy j, EZ «uebsy yryouura' io d wrexuor wn age

Toy juiv &yu VH ν,ju Eνu Hi.) 04 xanug A Tis abr, maurew odo, og p wi Vupuw Anpu Ine Qurv wa ovudec ronn' avbplurruv "Ng un Wergopovog , Ax,“ Kanon

Exb oppor uxeri wounay, &c.

The line Zevg ru xara&x0or0;, &c. is evidently a comment upon $0: in the preceeding one, and probably an improper comment; for, though Pluto and Proſerpine were before invoked as the deities of deſtruction, they were not peculiarly the impeders of generation. Neither does Homer in any other place call Pluto by this title ; which, being derived from the myſtic ſyſtem, was probably unknown to him. dad ros and acaToG;—ANAFATOE and ANAFAETOE, he regular 5jeo-

tives from the verb AFATQ; according to Dawes,

The N, however, to ſuſtain the privative A, though conſtantly uſed by later writers, does not regularly belong to Homer's or- thography (1); and as the £ was frequently elided, and the conſonant doubled, in the old dialects, the antient words were probably AAaFATOL and AAFATTOE, from which the change to the preſent reading was very eaſy. Heſychius has AAZTON, avapagryroNy aS, and AATON, with nearly the ſame explication ; but it 1s evident that an A has been loſt from both theſe words, and probably a T from the latter, as they are both the ſame, only formed according to different dialects. He adds, however, ano- ther explanation to the latter, ſignifying inſatiable, avernnwror, the reaſon for which will be given.

The omiſſion or inſertion of the ſubſidiary and paragogic N, having been left in a great meaſure to the diſcretion of tranſeribers, has, I believe, pro- duced conſiderable confuſion both in the meaning and etymology of ſeveral of Homer's words. Upon the medals of Alexandria Troas, the title of

(1) See II. 2. 536.


ON THE GREEK ALPHABET. _ Apollo, which we now write Zo, is un'formly EMIOETE, that is, in antient orthography, EMI@EFL, which has fo near a reſemblance to our word sMITE, and its various derivatives, that we cannot but ſuppoſe i it to have come from the ſame root, and to have ſigyified the $MITER or on- $TROYER, generally, according to a well-known attribute of Apollo, ex- preſſed in the ſymbolical writing of antient art by the bow and arrows

which he carried. The tale which deduces it from opnboy, faid to be the 

Cretan name for a mouſe, is of later times, and gives a ſignification unwor- thy of the ſolemnity of the occaſion on which Chryſes invokes the God, in his character of Deſtroyer, to avenge his wrongs upon the Greeks, Like many others of the ſame kind, it was invented to give a fictitious meaning to one of thoſe old myſtic titles, the real ſignification of which was kept concealed from the vulgar, Ariſtarchus rejected it, and derived the title from a city of the Troade (1), which appears, however, to have been un- known to Homer, and which was probably named-from the title. From an improper inſertion of the ſubfidiary N, as 1 am inclined to be - lieve, aroſe thoſe unaccountable forms of verbs avyvels and emervobey, which many have ſuppoſed to be præterites middle of avbew, to 5loom or bloſſom, with the Attic reduplication, and poetic inſertion of the O. But how there could have been a poetical licence of inſertion, when poetry was the only ſpecies of literary compoſition ; or how Homer's audience, who had no dictionaries and grammars to conſult, could have underſtood forms ſo remote from common uſe, I cannot conceive. The ſenſe alſo, as Dr. Clarke obſerved, requires an imperfe& rather than a perfect tenſe ; and the metaphor, according to this interpretation, is too forced and unnatural for Homer, who would ſcarcely have deſcribed the blood blooming from a wound (2), the fur blooming from, a ſkin (3), or the ſcattered hairs blooming upon a bald head (4). I cannot, therefore, but think that theſe words are compoſed of obe, 10 puſh or move, and are, therefore, regular imperfects ara- tobe and t- - toe, reduced by the ordinary craſis of the vowels to arnob and erevyobe, and then corrupted, by an —_— inſertion of the ſubſidiary N, to

(1) Apollon. Lex. ad Heſych. Albert. citat.

(2) II. A. 266. e

(3) II. k. 144. 5 = - (4) II. B. 219. . RE!

I 2 


axes and . The verb o indeed, does not Abbes occur in a neutral ſenſe; but moſt of the Greek verbs had a neutral as well as active and paſſive ſenſe, which is oftener expreſſed by the active than the middle voice(1), The pronoun might alſo, in theſe inſtances, be underſtood, as in o. jury avwoarre R tg mob (2). The Venetian Scholiaſt would, how- ever, derive -e and enwnols from «bw, antiently FEON, whence the perfect FEFOOA, now written u (3). According to his idea, therefore, the antient forms muſt have been ANFEFO®E and EIIENFEFOQE, or with the aſpirates elided, as in compounds, ANEFOOE and EHENEFOOE ; but the perfect tenſe will not do in either inſtance. In Od. ©. 365, the ſeuſe

ſeems indeed to favour this etymology ; but I think the line is ſpurious.

Erde d pur Kergirtg AUT ov Ka U tNoit (Alegre, Ga Jiu emevnrolev airy tv Aus J. Pear a doc ETMpaT o, Nau eo hi.

Au, or 40 —AFQ, or AFEQ ; whence come the antient words AFOL

-EFOE or -EOF, morning, and AFOFOE the adjective derived 

from it; both of which are now written and declined, after

the Tonic and Attic manner, HQE -OTE, and HOIOE (4). The Ionic va- riation might have taken place even before the time of Homer ; but the Attic termination of the genitive is, as before obſerved, a corruption of no very early date, it being unauthenticated by any very antient monument. An immenſe number of words are derived from this root, all of which were antiently written upon the ſame principle, as AFHP (in Ionic EFHP or HHP), AFHAQN, AFEIAN, AFIZEQ, &c, &c. The two laſt were contracted by the Attics to ade and goow; but theſe abbreviations could not have taken place whilſt the F was in uſe, wherefore they are unknown to Homer, who always makes the firſt ſyllable of aizow long. In the old editions, indeed, of the Hymn to Apollo we have &doy ( * ; but if this be

(1) Notum eſt omnia fere verba Græca activa, ſzpe & intranſitive 1 Damm. (2) Od. O. 552.

(3) Hence THOIA (that is FHOIA) n in Heſychius, 

(4) Other provincial forms are preſerved by Heſychius, as ABQ, T1, Aaxurg, and 5 ATC, Ma. i

(5) VI. 22. Clarke has ad, but cites no authority. 0


ON'THE GREEK ALPHABET, 61 the true reading (as I believe it is), it is an additional proof that this ele» gant poem is not Homer's, though quoted as genuine by Thueydides. Avy———ATFQ; whence are derived AYFOE, AYFTMH, ATFTH, &c.

- - It appears, however, from a paſſage of the Venetian Scholia, that Chryſippus the Stoick, and Dionyſius of Thrace, two antient Cri- ticks of great eminence, wrote aui % with the common aſpirate at.uxos, or AF TIAXO (1) ; conſiſtently with which, they muſt have written theſe words in the ſame manner, AFYQ, AFT TOS, &c. The authority of the beſt antient grammarians is, neyertheleſs, but little in the uſe of the aſpi- rates, and general analogy favours the F in this inſtance ; but, without the authority of monuments anterior to the ejection of theſe letters from the Alphabet, it is impoſſible to decide with certainty. Agg——FAPE. ;

Tore —— TD TFIEL, the plural of FrTFE, contraſted from PYFENE ; * whence, I believe, that aryvrie and amywreo are properly the ſame word, antiently written AITTFENEZ, the regular plural of AT TFEnx, a particular ſort of TTFEIZ, or vultur. Ale, bog, &c,——AFIQ, AFEOL, &c. according to Dawes. The vowels preceeding theſe words are uniformly long, whence the augments in the Aoriſt and perfe& tenſes have been

changed from E- and AE- to EA- and AEI-, as in 89%ox and de. Whether, however, the For the T was the letter that has been dropt, I have tome doubt, but am inclined to think the latter, for the word Zeus or AZEFZ, and the Latin Dsvs, are certainly yo this root (2) ; and that


(1) Ad IL N. 41.

(2) AETZ, Zevs, dioc, poCo;, 1 Sies, Heſych. The account of this title, in the new Syſtem of antient Mythology, is ſo new, and, at the ſame time, ſo comic, that it may ſerve to enlivert the dryneſs of the preſent Diſquiſition. Noah, according to the learned and in- .genious Author, not only planted vines, and made wine, to intoxicate himſelf, but like- wiſe ſowed barley, made malt, and brewed beer; which, being called in Greek Zubos, or

(as he chuſes to write it) Ziubos, became, though a very contemptible liquor among that people, the name of their ſupreme god ; who, it ſeems, was no other than Noah deified in the character of a great brewer,

Jupiter e quodcungue vides, guocungus moveris, ſays Cato, in Lucan ; bat, though Cato was fond of ſtrong drink, none but this ingenious gentleman, I believe, ever thought of giving 0 pleaſant a turn to his celebrated ſpeech, as the making him alleviate the real mi-



1 4

prides, An .


the L was oveificonlly dropt from the 4A, even in the. early times, is proved by the high authority of the Zanclèan medals before cited, and alſo by the names Zaxuvbog and Zea; which (as I have ventured to conclude from the facts above ſtated, and the analogy of the metre, which requires

a ſingle conſonant) were written, upon the ſame principle, AAKYNOOL


The uſe of the E, like that of the other aſpirates, depended much upon cuſtom or diale&; for though no licence could inſert it into a word to which it did not radically belong, it could, in almoſt any caſe, be elided. Hence the apparent irregularities in the oblique caſes of the word Zeus, which have, however, all been very naturally and regularly formed, from the different modes of writing and pronouncing it in different dialects, as

N. ALZEFL, ALHN, or ALIFZ, contracted, by eliding the aſpi- 5 rates to AlL. SG. AYEFOZ, AZHNOYL, or AZIFOZ, contracted to AIOE, D. AEZEFI, AEZHNI, or AIFF, contracted to ATI. A. AZEFA, ALHNA, or AEIFA, contracted to ATA. From the perfect tenſe of the verb ATI or AZEIN, the Greeks, as

uſual, formed new verbs, ſuch as dude, dude, &c. which ſhould pro- 

bably be written AEAEQ, AEAZILEN, &c. in Homer, the I having been apparently inſerted, as in many ather inſtances, to ſupply the place of the aſpirate,

In a very few inſtances out of the great number in which theſe words occur, the vowel preceeding is ſhort ; but this, I believe, is always owing to corruption. Ewmep adems Y «5+ (1) ſhould be EINEP T' (or K') EET” AAZEIHE. Bporryoag 3' ape devor (2), ſhould be BEONTHEAE AE AxEI- NON, the particle aa being unneceflary. The ſame alteration ſhould take place in 2 J'ape Jurarrur (3), and it may be generally obſerved through- out Homer, that the particles have been very licentiouſly employed by the antient editors and tranſcribers to fill the vacancies which a change of Al-

ſery of thirſt, which he felt upon the burning lands of Libya, with the ideal happineſs of

being immerged in a barrel of beer. ) II. H. 117. (2) II. 6. 133. {3) Odyſl. a. 533. phabet


phabet has produced in his metre. Ada: is, I believe, uſually pronounced in four ſyllables, the two firſt ſhort ; but it ought to be pronounced in three, the two firſt long, Arie or AEASCIAEL. :

The vowel having been thus uniformly long, is, I think, a further proof that the L was the letter joined to the A, and not the F, as Mr, Dawes ſuppoſed ; for there is no reaſon from analogy why the vowel. ſhould be always long before AF any more than before AF or . I am Kill more convinced of it, by finding the £ omitted in the flexion of a verb of ſimilar form, in the theme of which it is ſtill retained, Epidyoar- 6a, is evidently from te or EPIAT Ea, and ſhould therefore be regularly EPIALHEALOAI, though the claſhing of the rough and barbarous dentals induced either the Poet himſelf, or his antient editors, to prefer a trifling grammatical licence to, a harſhneſs of ſound. This licence, indeed, like every other employed by the Poet himſelf, appears to have been previouſly authoriſed by familiar uſe; for, as the true antient forms were probably EPIALQ and EPIAEELALOAT, which are conſiſtent with the other flexions of the ſame verb: the Tolians, who elided the aſpirates, aud doubled the conſonants, might have written and pronounced them EPIA- AQ and EPIAAELASOAI, the third ſyllable of the latter being rendered long by the emphaſis laid upon the E which terminates it. To prove that the T was occaſionally elided, and its place ſupplied by doubling the con- ſonant, we have alſo the authority of Plato, in whoſe Dialogue upon the Immortality of the Soul we find the Bœotian interlocutor employing ITTQ for Izra, which in Homer's time would have been, in that dia- let, FITTQ ; whence we may perceive the affinity between this verb and the Saxon pirzan, the root of our word wit. We likewiſe find, in the Lacedæmonian Decree againſt Timotheus beforementioned, AIAAKKE for EAIAAKEE, to which the Oxford Editor, with preſumptuous and inau- ſpicious hand, has changed it; not conſidering that Homer and Heſiod have employed repeatedly a fimilar form in a word which is now written Ine, according to the Ionian mode of extending the vowels and eliding the conſonants ; but which, in the old language, was QEKKE for /EOEKEE, the third perſon ſingular of the Aoriſt of Kn, the old olic form of E or Tien, it having been cuſtomary, in that diale&, to terminate verbs

in -KQ, which others terminated in- TT, -T EM or A -TTQ, and N pure:


pure: of which conſiderable remains are obſervable in the Dotic, and alſo in the future tenſes in -KEQ and -A of other dialects. Whether it was ever allowable to change the dental aſpirate for a dental mute in the begin- ning of a. word, and to write AAEINOE for AxEINOE, AAEIQ for AZEIN, &c. I cannot determine; but there is nothing in the analogy of the language againſt, it, and EAAEIEF, the third perſon ſingular of the Aoriſt, fo often repeated, and ſupported by the invariable teſtimony of ſo many manuſcripts and editions, is as great authority as there can be for any peculiarity of orthography not authenticated by antient ipſcriptions. Aog——AIFOE ; whence came the Latin pivus. The firſt ſyllable of this adjective is always long, whereas it is always ſhort in AIOE, the genitive of Alx, from which it is derived. Hence we have uniformly Jie, nobly-born, and Jimery; (properly STerrerys, as in Heſy- chius), Fove-deſcended ; the former having been antiently written AIFO- DENHE, and the latter AIEINETHE. Aow, contracted to iz——AOFQ, contracted to AQ, and varied by habi- tual or local corruptions to ATAQNMI, AOEKNQ, and AQFKQ ; from which laſt comes the Aoriſt dena, properly EAOFKKA for EAOFKZEA, often written without the aug- ment Juxx for AOFKKA, in the ſame manner as de for @EKKE, wee for FEKKE, &c, This cuſtom, however, of eliding the conſonants and aſpi- rates, and extending the vowels, being Ionic, might have taken place in the time of Homer, who, upon the ſame plan, has a. for IAT Al, ei- aaro for EDIAZLATO, &c. whence the Criticks have been much perplexed; for , notwithſtanding what Clarke ſays (1), has the firſt ſyllable al- ways ſhort. Eag—FEAP, written by Heſychius FEAP, according to his uſual prac- tice of putting the F for the F. Ea EF An, written by the Laconians and Syracuſians (who 1 in this

_ inſtance employed the Laconian dialect) EBA (2). 
Exrgrn—FEAIQ. The vowel being ſuſtained before this verb, proves 

that it began with an aſpirate; and I have been induced to prefer the b to the F by an jio(cription publiſhed by Abbe Winkelman,

(1) Ad. II. 1. 304. „„ Heſych, 


in which we find, in Latin letters, the Greek names MINDIA HELPIS (1),

the latter of which is evidently derived from this verb. Our word mELy

ſeems alſo to be of the ſame extraction; whence the verb To u was formerly declined nearly in the ſame manner as the Greek, HELP, HOLPEN


une and £61200 ;-——FIKOET and gFIXOZL. In the Heraclaan Inſcription

it is uniformly EIK ATI, except in one inſtance, where we

have EIKOZI, which is probably a miſtake of the graver for CIKOET, as EE TOR, in another inſtance, certainly is for TETOZ.

Eixo, eros, &c.——FEIIa, FEIIOE, &c. In Heſychius we have FINON

(that is FINION) «To ; but the ſubſtantive ſhews that

it ought to be written with the KE. 

Ed kla n. and FIAQ, with all the derivatives FEIAOx, AFIZE, AFI- AHE, &c. It appears from Heſychius, that the F was once pre- fixed to the A privative in ſuch words as the laſt ; whence he has

TAMMOPOE (that is FAMMOPO ZE) apogog; but this does not ſeem to

have been the orthography of Homer. Opp' 2,8% ſhould probably be o-

PA FIAQ, though the vowel may be elided before the F as well as be-

fore the P. TOLIAHMAI (that is FOIAHMAI) en:gapa: of Heſychius is

taken from the preterite of this verb, FEFOIAA, 3 written without the augment FOIAA.

.- is probably the original term, 6d rr the abbrevi-

ation; Lennep's doctrine of an adſcititious E, prefixed arbi-

trarily to certain words, being contrary to the analogy of every language; but eliding the firſt vowel was common in the Doric dialect, and probably in the old olic, from which it was derived. New forms alſo, both of verbs and nouns, aroſe from the DAE denſesy d and re- tained the additional ſyllable.

Edog, abs, abvog, & c. FEOSOE, FEONOL, * The ſyllable FE may an-

ſwer to the long and ſhort vowel, or the aſpirate 

might have been dropt occaſionally, and the vowel

extended, ſo that beg and yd are probably the ſame word, written differ-

ently according to the cuſtomary pronunciation of different countries, He- ſpychius has, however, THelA (that i is FHOIA) 205 but his authority in

) Hiſt. des Arts, I. IV. e. vii. T K the


the uſe of the double or ſingle yowels is very little. Mang tube ſhould

probably be MAAIETA FEFOk, the tranſmutation of the E into the O

being common in the perfect tenſe; and the O, in the prefent inſtance,

being rendered long by the aſpirates. Els may, however, poſſibly be an

Toniſm of the ſame kind as thoſe abovementioned. -

Ew, præt. 0 —FEIK n or FIKN, FEFOIKA, as Dawes has juſtly

obſerved; whence FEIKEAOZ, AFEIKQE or AFI-

KQE, FOIKOL, &c. Jes appears to be of the ſame

root, and accordingly we have PIT TON (that is FIEFON) e; and BIQNP as Laconian for ow; in Heſychius. To this the analogy of our word wise, in the compounds LIKEWISE, OTHERWISE, &c, exactly correſponds both in form and ſignification. The firſt ſyllable's being uniformly long too favours the orthography of Heſychius, as does likewiſe the regular pro- greſs of etymology—FIKOL, FIKEZOE—FIKEZEFOE contrafted to FIZ-

FOE. In the Heraclean tables, however (the only aſpirated inſcription in

which this word occurs) it is FIZOE ; but though authority is generally

to be preferred to analogy in matters of this kind, I think, in this inſtance, we may ſafely attribute the peculiarity to local corruption.

EF ETO: wherefore the firſt ſyllable is frequently long and the ſecond ſhort, - Barnes, indeed, ſuppoſed that ig «yw, at the be- ginning of a line, was an amphibrachys, equal to a dactyle; and

Clarke, ſtill more abſurdly, that it ought to be pronounced as a ſpondee,

by a ſort of metatheſis, woe % (1). The learned author of the book upon

Rhythm would, in one place, divide the intermediate long ſyllable in a

manner which I avow myſelf incapable of exactly comprehending (2); and, in another, elide the firſt ſyllable (3), as the Dorians frequently did; but, nevertheleſs, without extending the third in conſequence of it, as he muſt do to fill the metre... All theſe refined conjeQures are, however, ſu-

perfluous, if we read the word in its original form and antient letters, 

ſome paſſages, indeed, we find it in one ſyllable, as T' deg fue g ue (4). 2 [uy 0 0 de TW; uu e pEYaporg PUARNELO (5); : Eu juw (al. f) i o =-


(1) Od. 4. 120. (2) Lib. fing. de Ryth. 8 P- 37. (3) Ibid. p. 142. (4) Od. B. 148. 5) 0. 0. 131. (6) Od. E. 123.


ON THE GREEK ALPHABET. 67 Eos he vag re deu (i). But in each of theſe there is ſomething redundant. In the two firſt the particle e encumbers the ſenſe as well as the metre; aud, in the third, the pronoun ſhould be changed from uw to the old regular form 6—-EFOS -* EN OPTTTIHI. The fourth has been corrupted by two different read- ings, ev and yep being (as has frequently happened) joined in the text, the firſt of which is, in this inſtance, the beſt—FEFOE MEN TE @EFOYEI— In another paſſage of the Odyfley, wg ore is written for FEEFOL—Heodbs

& og br aodeg en f, acide (2), inſtead of HEOIE A' FEFOE AFOIAOE

ENI MEPAPOILILIN AFEIAEN ; and though Bentley found zg in a manu- ſcript, Clarke did not chuſe to adopt it, becauſe Euſtathius and the Scho- liaſt have dc are. With the ſame timidity or negligence, and equally to the detriment of the ſenſe, he has preſerved KAI for KEN, in Odyfi. P. 146, though the true reading is retained in A. 560, where the ſame line occurs, Ou ya d Wope vits nee xau Ira Ol Kev july jj 67" eupecs YH, N-

Huxp——FHMAP probably, like ie or FHMEPA, whence the vowel is often open before it.

Duo, &c. er Fa, &c. the firſt ſyllable being 3 long. Iaxw—FIFAXQ according to Dawes ; but it ſhould rather be FIAXQ, for the firſt F 1s ſufficient to prolong the ſyllable and ſuſtain the preceeding vowel, and there is no authority or reaſon for inſert- ing the ſecond. His emendation of au@iaxuar to FEFIFAXYIFAN has certainly produced a much more monſtrous word than any he could have found to remove; for ſuch a flexion as -YIFA from · Ax, or indeed from any other termination, could not have exiſted at any period or in any dia- le&, it being inconſiſtent with the analogy of the language. If he had thought -YIA not ſufficiently archaiic, he might have propofed - TFA or -OFA, which, though unſupported by authority, agree with the general principle of declination. The preſent reading AMSIAXTIAN is, however, probably right; the omiſſion of the augment being common, and the eli- fion of the aſpirates in compound words juſtified, not only by the frail ſyſ- tems of the antient grammarians and ſcholiaſts, but by the indiſputable

t) U. LY EIT e K 2 authority


7 © »


authority of the Veletrian Inſcription, in which the word FOIKIA is writ- ten with the Digamma, whilſt AAMIOPTOE (which according to etymo- logy ſhould be AAMIFOPTOE) is without it (1). The prepoſition AM®1I has alſo a peculiar beauty in expreſſing the tenderneſs of the mother flut- tering round her plundered neſt whilecrying out. Hxy and qx« are only variations from the ſame root, and therefore were written FHXH and


I3,07=——LIAIOE uniformly i in the Heraelèan inſcriptions ; ; but the metre does not require the aſpirate in any of the inſtances where Homer employs this word,

1— In.

Iergy ibo, &c.— FIPIEZ, FIpOx, &c.

Ig, b, &. — FIZ, Flo, &c.; whence comes the Latin vis, and the TIZK TN and BIEXTN, both explained ;oyv, of Heſy- chius. From the ſame root are probably derived wp: and

TIL which ſhould therefore begin equally with the F.

Ira—FITEA; whence it is TITEA in Heſychius.


KTamuai ———KTAFOMAT. Hence x7TyoaTo ow; in Oy. F. 450, ſhould be

KTAFEZAT' OIOF ; this being the old form of the Aoriſt, as

I ſhall ſhow in conſidering the flexions of the verbs. Kryrig

ſhould alſo be KTAFEIE or KTEF ZIL, from the Tonic form KTEFOMAI1, whence KTEFMA, now xTyua, KTEFAYZ, KTEFATITEQ, now xTe&-

rig, &c.

Kuayg, &c,—KYFANEOE, &c. whence the firſt ſyllable 3 1s "=

KuJog, &c.—kKTFAOZ, &c. probably from the ſame root.

Kvpa—— KTFMA.

Konus, xwxurog———KANKTFAN, KNKTFTOE.

Aauw——AAFQ, written, through a difference of dialect, AAB; which,

acquiring a metaphorical meaning, became a different word, as it uniformly 1s in Homer. The derivatives ſhould all be written after the ſame manner, which will be found equally conformable to the

rules of metre and etymology ; as AAFOZ, AAFAE or AAFZ, AAFINT E,

(1) In the Heraclian inſcriptions the aſpirate is uſually retained in the compounds.

2   AAFPH, 


AAFPH, ATIOAAFPA, &c.; > alſo the proper names from theſe. roots, ſuch

as AAFEPTHE, AAFOAAMAE,, AEFKOQEFH, AEFKINOOE, &c. The original verb ſeems to have been antiently written with the F, employed as a guttural aſpirate to expreſs the rough pronunciation of the old Æolian and Pelaſgian clans; FAAFQ, whence yaaurow or TAAFEEQ, which is only a different mode of pronouncing AEF ZT. This gives us the true etymology and ſiguification of yAavzwr; or TAAFK nx, the epithet of Minerva, which means neither blue-eyed nor owl-eyed, but keen-eyed or eager- eyed, having extremely quick and comprehenſive fight, as Heſychius has rightly explained it. DTAAFKE, an owl, was ſo called from this quality; and TAAFKOE, the adjective, ſignifies the activity and violence of the ſea rather than any particular colour; whence FAAFKIOF N is employed as the epithet of a lion darting upon his enemy, to expreſs the eagerneſs and ferocity of his look (1).

Acuo —" AOFQ. Hence Avoca; and Mugen, which are the ſame forms of

the Aoriſt, except that the one is contracted and the other not; AOFLAZE and AOFEEAYZ, the penultimate E of which may be

pronounced, as uſual, double or ſingle. Abo — ATF. The aſpirate is elided in ſome of the flexions, and alſo in the adjective and abſtract ſubſtantive derived from it, AT TOx and AT IZ. This ſeeming irregularity perplexed M. L' Abbe very

much; but Dr. Clarke treats his doubts with ſome contempt; and, to pre-

vent any one elſe from doubting, aſſures us, with great gravity, that it was an eſtabliſhed cuſtom to pronounce the penultimate ſhort in ſome flexions and derivations of the ſame words, and long in others (2). Of

this L'Abbe had certainly no doubt, as the knowing it was the only,

ground upon which he could enquire into the-cauſe of its being ſo. It did not, it ſeems, occur to him, that exactly the fame kind of eliſion takes place in the flexions of ſome Latin verbs, as Aubir, PERIT, and vu, far

AUDIVI1, PERIVI, and FUvi, where the correſpondent letter to the Digam-

ma is funk. Both Clarke and Barnes ſuppoſe Avro tp be au abbreviation of 8a8Auro, otherwiſe, they ſay, the T would be long (3). But this is a law of their own enacting; for the aſpirate might be as n and properly

(1) See II. 1. 172, and 11. 34; and Schol, Ven. (2) In Il. A. 314 (3) II. o. Vſ. 114; and n. vt. 1.

1 ” —_ 0 1 * — bo ©

T . - * .


"+ "ab


elided in the imperfect (of which the ſecond Aoriſt is a particular form) as in the perfect tenſe; and, in ſome inſtances where hure is uſed, the ſenſe will not admit of a paſt-perfe& without confounding and perplexing the narrative of the cleareſt and moſt accurate narrator that ever wrote. It is in theſe ſecond Aoriſts too that the principal eliſions take place through the flexions of all the verbs, as eruroy from TumTw, «Gao from GH

h from pa, &c.

NM MAFA. Hence weucoros and euros 2 MEMAFOTOS, the genitive of the participle MEMAFQYE, the F being, as uſual, ſounded with*either vowel, Muarzy ſeems to be an abbreviation


(fee Damm. Lex, Etym.). It may, however, be the Aoriſt of a new

theme from the perfect.


O;;— —OFIE, whence the Latin ovis. In the oblique caſes it is often

pronounced in two long ſyllables, and often in a long and ſhort one, as vg awry, Which, unleſs the F was elided, muſt have been pronounced OF-IOZ AFT g, as it might have been without any violation of the laws of profody; for, as OI and EI are ſometimes ſhort in OIOE and ENEIH, 10 might be equally ſo in the preſent caſe. The F might, however, have been occaſionally elided as well as the A, the regular flexion being OFIE OFIAOE, The accuſative plural is, in the preſent text of Ho- mer, contracted to o, with the firſt ſyllable ſhort; but, as the ſecond is always long, it might antiently have been written and pronounced regu- larly GFIAx.

Owog——FOINOE; whence, through he medium of the Late: our word

wine. Heſychius has, as uſual, LOINOE—OINOE,

Owe FOMoa, it being derived from FOMOE.

INivv, &c.— lf; whence NIFAP, IIFAAKE, &c.

n —— MNOAIE lor m -IN -IEZ -IQN -IEZI -IAx.

IOAEFE -EFOZ -EFI -EFA -EFEZ «EF NN »-EFEEI -EFAE, now written = -y "1% eg =1wy -8U0s "1&5. TNexirys or TIOAIF=

"THE ſeems formed out of both, unleſs it was antiently written NOATE-=

THE, which uſually ſignifies the founder of a city; but in the Heraclean

Tablet we have the genitive plural NOAIZTQN, ſiguifying the ordinary

4 inhabitants.


inhabitants. The datives woot, dh, 8c: are probably from fimildr ob» ſolete forms, NOLEFE, AFLAPEFE, &c. and not, as is generally ſup- poſed, Ionic flexions of the common terminations in -IE, mus or wheiw——TAEFQ or TAEIA ; whence ITAEFIAEZ, now te, +, +», © and NAEITAAEZ, the plurals of two different forms of the ſame word AE FIA and HAEIAZ. . II IT TF; hence the ſubſtantive T C. n or ruhen or nren; whence IITFOE, IITFEAOY, &c. the- future of this verb, uo, ſeems to be formed from the firſt theme, 'ITFEQ, unleſs indeed it be formed by eliſion of the ©, as oow from oh. Erynneeg——EIFAAOFEIE according to Dr. Taylor, who decides it to be the participle of the verb EIAAQEAT, S of Heſychius, the theme of which he of courſe ſuppoſes to have been ZIFA- AOFQ (1 ). Heſychius, however, ſays alſo, that the material employed by curriers to prepare leather was called ZIPAAQMA ; wherefore, as Hemſterhuiſe has obſerved, the preſent orthography muſt be right, unleſs (as has frequently happened in Heſychius, but never, that. , know of, in Homer) the F was changed to a T. Fe! - Dries IE FOx or ENEIOE. Hence the 4 plural nei for .- FNEFEEZEI or ENEFEZT, and oreoo: for the contracted form SIIEFLI. The genitive, according to the uſual change, has be- come cνε, Which may, however, be read ZITEFEOL or ENEIEOS, in every inſtance, without injuring the metre. Tie, Tice, &c,——TIFQ, TIFEQ, xc. Hence the firſt {ſyllable in the. future, &. is always long, though ſhort, as uſual, by the eliſiou of the aſpirate, in the abſtract ſubſtantive and


'T Uy as Trp ——TPYDAFQ, TPYPOFQ; whence TPYDOFQEIN, and TPYDOFOIEN 3 by elifion of the F, TPYDOOIEN,

now: written regen. ; This verb ſiguifies the action of  

in, or depriving any thing of us run ; ; whence ATPYTETOZ has

8 (1) Let. Lyſ. C. IX.  


been ſuppoſed to mean ferile, that from which no fruit can be gathered, or, which is totally anproduttive. It is, however, applied to water and æther, the elements which are generally characteriſed as the ſource of all production (Han ran your, and yiverug); Wherefore, I am inclined to think that it

means hat whith in ſo productive that it cannot be exhauſted or deprived of

its produce, The aſpirate being dropt from the verbs of this form has given an appearance of licence in ſome of the flexions where there 1s really none. Thus we find 94\cwvre; and yerworre;, both of which are properly the ſame, TEAOFONTEZ, the F being pronounced equally with either O. 1 Tra; whence FEYFAAEE, FYFETOE, &c. Yay ——FTAFH. In the old Folian EYAFA, whence the Latin s8YLva. ou, Prog, Prov, &. 0 rFa, OYFKOEL, orFAON, &c. In the ad- jective, formed as uſual from the ſecond Aoriſt, or contracted imperfect of the verb, the aſpirate is elided oöreg; whence a new verb, qurevw or OTTEFQ, was formed, which the Latins adopted in an obſcene ſenſe. On an antient baſe of a ſtatue, in the iſland of Delos, we have O AFYTO Aleo EMI ANAPIAL KAI TO roEAAZ; which Dawes would correct to TO AFTO AleO EMI O ANAPIAT TE KAI TO E®EAAZ, the article having been, as he ſuppoſes, mutilated in the firſt inſtance, and omitted in the ſecond, through a blunder of the tranſcriber, and the T inſerted in the pronoun by a later hand, as a comment upon the F. Both theſe conjectures are, how- ever, very improbable; and I have been aſſured by thoſe who have ſeen the ſtone, that the letters are preciſely as in the annexed plate (1). Is it not poſſible that the article at the beginning may, by a local ſingularity of ſyntax, refer to ANAPIAE, and AFTTO ſtand for FAT TO, ſignifying the ſame as FO AYTO, of which it is a coutraction? for, though there cannot be any arbitrary tranſpoſition of letters in any language, ſuch corruptions

might eaſily ariſe amidſt the licentious variations of local habits in a lan-

guage which had no fixed rules of eſtabliſhed practice to confine it to ety- mology. In the ſame manner, therefore, as AFIAHE became EAAHE by a corrupt local change and tranſpoſition” of the aſpirates, FAYTOL might

have become AFTTOL, which might have been equally pronounced in 

(3) Plate I. Fig. 3. 785



two ſyllables; for Fr might have been pronounced merely as an empha- tical W, it being always to be remembered, that neither of the two vowel aſpirates ſiguify, of themſelves, either tone or articulation, but merely certain modifications of them. Hence Homer has ov anc; (chat is LE F“ AAAO;z) in two ſyllables (1); and Virgil bantxe in one (29.

numerary character arbitrarily inſerted, in a public inſcription conſiſting of ſeo few words, and exhibited during ſo many ages in one of the moſt cele- brated and frequented ſpots of the antient world, I cannot admit, and muſt therefore think it inexplicable if it cannot be explained without alteration. II. When we find a long or double vowel; where ety mology can account for only a ſingle one, it will, I believe, invariably appear, upon analyſing the word, that ſuch double vowel has been introduced merely to fill the vacuity in the metre cauſed by the omiſſion of the aſpirate, which will be found as requiſite to give the word. its regular ſtructure as its proper quantity.

Some inſtances of this have been already cited, and many others will pre-

ſent themſelves to the attentive readers of Homer, as

Beirng——BPIEEFIE., Alſo other patrony mies and adjectives of the ſame claſs, as XPYLEFIZ, XAOPEFIL, &c.

Dprv; and yprug—— DPEFEL and DPEFTE.

Hire=———EFTTE. It occurs once as a diſſyllable ©); but Ariſtarchus diſ- 

covered that this was corrupt, and therefore, in his firſt edition,

changed it to EYTE.' Upon more mature confideration, how» 

ever, he found that aur was the true reading, which he judiciouſly ſubſti-

tuted in his ſecond (4), and which has been happily retrieved from oblivion by M. de Villoiſon's important diſcovery in the library of St. Mark's, at Venice, to the great improvement of one of the fineſt paſſages i in Homer :

11 6h Tio aury & ores Treg Axis 

Ei 6: 8@appeortity nai rrerx;o1 oyhac Wi, Te Naur rep yiver” ape bn Womnrm Keauv.

io avre, in the firſt line, ſhould be AE FEF' AYTOF, to make the eli- fon *

(a) u n „ 4) &n. IX. 40.

(3) U. T. 386 {4) Schol. Ven. in Loc.

That there could be any literal error of ſo much importance, or ſuper« 


Kagna, nan KAP EFA, KAPEF', or K APH; generally conſidered as an anomalous and indeclinable word ; but it appears really to be an abbreviation by Apocope of KAPEFAZ, written by

the Folians KAPAFAE, and thence contracted to KPAFAE and KPAFE,

now written ag and xpxg ; whence comes the verb KPAFQ, pronounced by the Ionians KPEIQ, a verb ſignifying ſupremacy and command, of which the participle KPEIQN only ſeems to have been in uſe in Homer's time.

Kuag and xwog——— KOFAL and KkOFOE.

Aus, e, c. AE FIE, AEFION, &c. probably h AEF a, the Cs verb as AABQ, written in a different dialect; ſuch

changes being, as before obſerved,” extremely common.

Hence we find both aygog and eg, which are the ſame word, antiently

written AEFIETOE, and pronounced with the firſt ern either loug or

ſhort, as ſuited the purpoſe of the writer...

Marnier MANTEFIOE, from MANTE Fa, or 8 cher adjec-

tives in -s, and ſubſtantives in -e, were formed upon the ſame plan, and conſequently written in the ſame manner.

Myrpwog, pyTpwog——-—MHTPOFOE, MHTPOFIOE, The firſt ſyllable of

theſe words was probably extended originally by the

aſpirate, and not by the double vowel; for uyryp is derived from Mara, and therefore was written MAFTHP, till adapted to the Ionic pronunciation MEFTHP-or MHTHP. For this reaſon. the firſt ſyllable is always long, while that of NATHP is ſhort, it being derived from nan. and not from A now written Wavws which, has a differs

ent and incompatible meaning. 15

Ilnog——TEFOE, or rather HA FOE, 8 » to the more antient Ko-

Mn lian pronunciation.

Newv——TIOFT; or, perhaps, IIAOFT or. NAOT, pally Aa; the june- 

tion of the A and O in an Q being common,

Pyidiog, pyigogy &. —PEFIAIOFE, PEFILTOF, &c. from PEFA, written 
; _ PEIA, in the Ionic manner, as oſten as the firſt ſyl- 

lable is pronounced long. In one inſtance only it occurs as a ſingle ſyllable at the end of a line—s ps Hg (1); where

(a) In: 101. f the


the aſpirate muſt have been elided, unleſs, as I ſuſpect, the paſſage be cor- rupt. The Venetian Manuſcript has u ue H · ; from which, com- pared with the other, I think the true reading may bo diſcovered=OT KE ME PEFA or PEIA. + Pa is, indeed, pronounced in one ſyllable in another paſſage (1), according to the preſent reading; whence Barnes, upon the authority of a. manuſcript, altered it to ia. Pama, however, is twice pro- nounced as one ſyllable in the Proœm to Heſiod's 4% uni jpupar—as pra pty tyap Perot, pee Is Gp % T ν, e, and though this Prom, conſiſt- ing of the firſt ten lines, be the contemptible performance of ſome rhapſo- diſt, it is, nevertheleſs, of ſufficient antiquity to prove that the antient copies of Homer exhibited the paſſage in queſtion in the ſame form as we now have it. I cannot, however, but think that it is erroneous, aud that inſtead of p7ia piv yap, we ſhould read PE IA MEN AP; the latter particle being much better adapted to the ſenſe as well as the metre, than the for- mer; for the reference is not to the ſimile of the vultur, contained in the preceeding line, but to the en action of Automedon ec ab in that


Too) Fr angel, rer „ &Xmupueveg rf 87 %ips u InTog argu, wo" GUYUTIOG [METH N. PEIA.MEN AP @wy«0xt un” 8% Tgwiy opuparydu,

_ Peics d emrcui6:0%6g \Wou N dH, omatur 17 hte 

Tyitr0g—TEFITETOE or TAFTTETOE, probably Ae! from TE-

FOE or TAFOE, though its being the proper name of a DJ 

mountain renders the etymology leſs certain, there being no W information to be had from the ſenſe. III. The ſubſidiary I and 1. which, Euſtathius ſays, the early Greek 7.3 91 writers yery generally affixed to the K and O (2), have very often ſupplied 4 22 14 . the vacuity cauſed by the Joſs of the Digamma as well as the double vow- el _ els. Hence we have, in different dialects, MOYEA, MOIEA, and MQEA, LY 7 4-6 5 whilſt the antient form was MOFA, from the obſolete verb MOFQ, from which came the Latin Moveo. The Laconians elided the Z, and wrote MOA; or perhaps, in earlier times, MOFA (3) or MOBA. - Kpuveg is alſo from KOF a, and therefore ſhould be written KPOFNOZE ; but, neverthe-

(1) 11.P. 461. (2) P. 511. { 3) See Decree againſt Timotheus.

L 2 leſs,


leſs, it is impoſſible to decide whether the practice of Homer's age and country was, in theſe reſpecta, ſtrictly conformable to etymology, or whether local habits had not even then changed the aſpitate to a vowel in many inſtances, Dawes would write AXEAQFOE for AXEAQIOE ; and there is no doubt but that, in the Dorian and ZEolian countries, the name of the river was ſo pronounced ; but Homer might have pronounced it differently, as an Ionian, as be appears to have done in the name of the city Elis, which, though beginning with the F on the medals, ſeems to have begun with the vowel in all the numerous inſtances where he men- tions it. AXEAQIOE might alſo have been pronounced in three ſyllables as well as AO al in two, though it might likewiſe have been pronounced in four, In the ordinary manner of writing it Axes, the Iota is equally retained though placed under'the preceeding vowel, according to the mode adopted in the manuſcripts of the middle ages, inſtead of after it, accord- ing to that of all antient inſcriptions. It is probable that the termination of the adjectives in -YE and -HE was once in -EFE or -IFZ, eontracted from -EFOE or -IFOE, the -1vus of the Latins;-and that thence came the formation of the feminine in -EFA or -IFA, now written -EIA. The tranſition from E to I is extremely eaſy, ſo that -EFA or -IFA might have been only variations of dialects. The termination in -IOZ belongs to a different claſs, and anſwers to the Latin in -1vs, the penultimate of which is always ſhort in both languages, whereas it is always long in -1vvs,



OY TO OO HETHER Homer's pronunciation and orthography, which muſt

have been thoſe of his age and country, were moſt Tonic or Molic, it is impoſſible for us now to aſcertain, though general tradition, and the

preſent ſtate of his text, join in favour of Ioniſm, At all events, it is ſafeſt

to ſuppoſe our preſent copies right, unleſs where anomalies, or ambiguities of metre or grammar, certain analogy, or antient authority, prove them to be wrong. As the removal of the anomalies and ambiguities can, in al-

moſt all inſtances, be accompliſhed merely by reſtoring the antient Alpha- bet and orthography, without ever changing the ſenſe, and ſcarcely ever

the order of the words, we may conclude that our text is, upon the whole, accurate. The tones and flexions have, indeed, been changed, as local or

temporary faſhion required ; but in other reſpects, I believe, the poet paſſed

through the hands of his Athenian and Alexandrine editors with leſs muti-

lation and injury than Shakeſpeare ſuffered from their ſucceſſors at Oxford. The very learned Dr. Heyne, indeed, thinks that, as far as relates to the

integrity of the ſenſe, he has ſuffered leſs than Virgil and' Horace, and many other writers, both Greek and Latin, who have flouriſhed ſince the Chriſ-

tian æra; and I cannot but think this opinion right, though Wolfius and Villoiſon have employed many learned and ingenious arguments to prove

the contrary (1). The latter in particular has given a curious and elaborate account of the various editions through which his works paſt, and the vo- luminous difputes of the criticks concerning the right readings, the ambi- guity and obſcurity of which was a-very antient ſubje& of complaint, It appears, however, from the ſpecimens of them publiſhed in his Venetian Scholia, that their diſputes were in general minute and frivolous, and the amendments they propoſed ſeldom well-judged ; ſo that all perſons of real

(1) Wolf, Praf, in Heſiod. Theog. Villoiſon. Prolegom. in Homer, 3 taſte


taſte and diſcernment among the antients ſought for the old editions which had never paſſed through their hands. Interpolations there certainly were, and ſtill are, in the text; but even thoſe conſiſt chiefly of verſes which are really of the Poet's own compoſing, but which the Homeriſts choſe to re- pcat in places where he did not intend them to be introduced. The lines that are rcally ſpurious are principally marginal explanations which have

ſlipped into the text, where they are generally ſo eaſily diſcernible, that we caunot but wonder how the Alexandrine Criticks could have overlooked

them, more eſpecially when we conſider that their extreme faſtidiouſneſs induced them to condemn verſes undoubtedly genuine. Shall we ſay with. an ingenious gentleman fond of paradoxes, that the Greeks did not under- ſtand their own language (1)? Without going ſo far, we may venture to affirm, that the writers who ſucceedeg the Macedonian Conqueſt, and conſidered the later Attic as the univerſal dialect, and ſtandard for purity, were not likely to form very accurate notions of the ſtyle of Homer; for, inſtead of conſidering their own grammatical flexious as corruptions of his, they conſidered his as licentious or poetical deviations from their own ; . wherefore they began their reſearches at the wrong end, and conſequently, the farther they purſued them the farther they were from the truth,

Happily, however, Homer appears to have had a Steevens among his editors, as well as many Hanmers and Warburtons ; for our preſent co- pies are certainly leſs adulterated than thoſe which were read by the moſt learned of the Attic, and later Hellenic, writers. Many, indeed, of the citations which we find ſcattered through the works of the orators, hiſto- rians, and philoſophers, might have been incorrectly quoted from memory, or corrupted by tranſcribers ; but others are ſo remote from the preſent reading, that they muſt have been taken from different copies. We find, for inſtance, both in Plato and Plutarch (2), the 528th ale of the laſt lliad cited:

Kypuy Suse c 250 0, auTup 9 dN, 

Inſtesd of

Awpuy, d 1 c, KaKkwys U 0s Env; iich is ſo n that the quotation muſt have been from ſome reverſi-

(1) See new Syſtem of antient Mythology. (2) Plat. de Repub, lib, i]. ; Plutarch. de aud. Poet.


ON THE GREEK ALPHABET. 59 fication into-more modern diale& ; for the uſe of the word inp, as em- ployed in it, is not of Homer's age. It may, indeed, appear preſumptuous in a modern Critick to diſſent from Plato and Plutarch concerning the right uſe of a Greek word, or the right reading of a verſe in a Greek poet ; 7 but in this inſtance we have ſtill greater authority to oppoſe to them in

ſupport of critical analogy. Vindar, who lived a century before Plato, and 

who probably read Homer-in his own diale& before he had been newly dreſſed by the Athenian editors, alludes to the paſſage in words which clearly prove that he read it as we now have it, though he underſtood it in a ſenſe ſomewhat different from the common interpretation, which ap- pears nevertheleſs to be right.;

Er wap 60h, r our

Avo d aloy ra Gporoig 

Abavaro (1).

But though the general ſenſe of Homer has been reſeued from N tion, it does not follow that the minuter accuracies of his language have not been extremely corrupted. Beſides the changes in the flexions and or- thography, the articles, particles, and prepoſitions (in the uſe of which the antient Zolic and Tonic Greek differed extremely from the Attic), have been frequently omitted, transferred, and inſerted, much to the detriment. of the metre, and critical nicety of the expreſſion, though I do not recol- lect more than one inſtance in which the general ſenſe is injured, fo as to make the corruption diſcernable to any but a very experienced eye, This is in the XXIId Iliad, where Hector, certain of his death, on finding. him» ſelf oppoſed, unaſſiſted and alone, to Achilles, 5000

Nvy 6 0 &yyub: poor Feeverrog xaxog, M de r aveuber,

Ou?” any” 1 yap pa War To ye PATH nav

Zvi Te ou Avog diet ExnConw, 0 uh Wap YE

TIpo@poveg e1guaTHai* YU GUTE j48 pOIpt KIN GVEL,  

Which, in its preſent form, literally ſignifies Evil death is near neut

even ſeparate——nor refuge—for it was indeed formerly agreeable to Jupiter and Apollo, who before cordially defended me; but now Pate overtakes me.

(1) Pyth. III. 145.  . 

b Inſtead



Inſtead of which, by only dropping the nn from the n and dem e particle, we have

1 jabs Darercy nance & er” n

' Oux ain tyoup'——y p rare Toys PirTIpor nev, &c. Evil death is near me not even ſeparate ; for no refuge. It wat, indeed, formerly agreeable to Jupiter and Apollo, &c. &c. The break in the ſen-

tence after ye, where r «51 is underſtood, has peculiar beauty in ex- preſſing the troubled ſtate of Hefor's mind; but, as the antient copies had

no points or marks to diſtinguiſh it, the paſſage was miſunderſtood, and

then corrupted to give it another meaning, or rather no meaning. The languid uniformity of modern language, which requires a continual repe- tition of the verbs and pronouns to make it intelligible, is ſcarcely ſuſcep-

tible of this beauty.

In many other paſſages of the Nliad and Odyſſey theſe minuter parts of ſpeech have been equally deranged, of which there needs no other proof, than that all the manuſcripts and old editions, which have hitherto been

inſpected, differ in a variety of inſtances ; nor is there one from which

ſome emendation has not been drawn. Much, however, yet remains to

be done—more perhaps than can be done with ſuch aids as we are likely to have; for though the ſtri adherence to analogy, which characteriſes Ho- mer's language, may guide us to the true form of his words in general, the almoſt imperceptible nicety with which theſe indeclinable particles

were uſed in the ſtructure and connexion of the ſentences, renders it fre-

quently impoſſible to decide where they might or might not have been in- troduced, The tranſcribers, however, oftener tranſgreſſed in the omiſſion than the inſertion of them, as appears from the various readings collected from manuſcripts and old editions; and I am perſuaded rhat, could we re- call them all, with the aſpirates, into their proper places, it would be found that all the arbitrary extenſions or ſuſtentations of the vowels, by the

,cſure, or otherwiſe, as well as all other anomalies, en, and ob-

"Doririgs, would diſappear.

When I ſpeak of ambiguities and obſeurities i in Homer, I do not mean ambiguities and obſcurities of ſenſe ſo much as of form; for the luminous

  • of his ſtyle is ſuch, that his meaning is almoſt always clear and

5 obvious ;


obvious; though, by the omiſſion of the aſpirates E and F, and other

changes in the diale& and orthography, many of his words have loſt their difference of form; whilſt they have retained their difference of meaning.

This will appear diſtinctly by comparing and examining the W words in their avtient and modern forms:

I. 1. A duco Arg. 2. Ayu fraugo FAT g.

The firſt was probably once written FATQ, whence the Ionic form FECEQ ; but Homer ſeems to have uſed it without the aſpitate, though he often employs the olic form Ar Ara, derived from the paſt tenſes.

The ſecond was written by the Laconians BAT, as appears from the explanation of BAZON and BATOE in Heſychius (1); whence we may _ conclude that it was generally written as here propoſed, It may be ob-

ſerved too, that the, augment, which always coaleſces with the initial of the firſt, ſo as to make , is always detached from the ſecond in Homer,

fo as to make «yo»; the place which the aſpirate occupied in the antient form EFATON, being kept void by the metre, We have indeed, in one inſtance, an exception—inT«uo ds d N Evyor; but, by only changing the eliſion, we may make it regular—HIIINEION A FOIL EFADLE GEA AETTON.

In two paſſages of Heſiod, and no where elſe that I know of, we have the ſingular word xave;a;; employed in the ſame ſenſe; which, I there- fore conclude, was compoſed of this verb and the n. ara, con- tracted, as it often was in compoſition:

— — 7 x

h ——äü'ͥ ——


KAFADEAIZ on — &@Z ove. xaeuaF ag ArZONA KAFAPEAIE (2).

The a ee which are numerous from both theſe verbs, of carlo

follow the roots, though now written, in many inſtances, without any variation, and only diſcriminated by the ſenſe. a ATH or AFA, admiratio (from which are formed the verb APAOMAL, ſhortened by eliſion to A AM Al, and the adjective Ar Aro; or eyaus;) be · longs to a different root from either.

(3) Ed. Alberi. (a) 57 L 1. VI. 611 & 638, ed. Brunk.

7 * Ma Cold. l. ten * 6

II. 1.



II. 1. 4 places FAAQ, 2. ad of alw' ſatio FAAQ or FAAEQ,

From the firſt comes «a3ora, properly FEFAAOTA ; or, without the Ionic cliſion of the charaQteriſtic letter of the perfect tenſe, FEE AAKOTA, the accuſative ſingular of the participle perfect of FAAaQ; which Dawes would write FEFANAOTA, but without any good authority, The editors and commentators of Heſychius would, indeed, perſuade us that FAAEIN, FAAELOAL, FANAANEIN, &c. which appear evidently, from the expla- nations annexed to them, to be forms of this verb, are literal errors, com- mon in that author, for FAAEIN, FAAEL AI, FANAANEIN, &c.; but the analogy of TANYMAI, Theka, and the Latin GAvpEo, prove that this was the antient and primitive form, anterior to Homer. The I was, as before obſerved, occaſionally employed as a guttural aſpirate, which, in the progreſs of refinement, was ſoftened down to the F, not only in this, but in other inſtances, as in PEPET or CxREs, originally the ſame name as that which in more poliſhed dialect was written EHPH, and applied to another perſonification of the ſame deity. I have, therefore, no doubt but that the preſeut mode of aſpirating this verb is right according to the lo-

nian dialect of Homer; and that, conſequently, 1/3ave ſhould be E-AN- AAN, or without the augment, FANAANE ; the ſpecies of metatheſis employed in the preſent orthography being, I believe, like every other, a corruption of later times, unkuown to the pure and regular dition of Ho- mer. At preſent, indeed, the ſimple aſpirate is uſually transferred from the theme of the verb to the augment; whence we have uniformly ids for EFAAE, which an autient ſcholiaſt very gravely tells us was a diæreſis of the corrupt contraction #4 (1). This is perfectly conſiſtent with the ab- ſurd prejudices of the old grammarians concerning the purity of the Attic dialect, but as inconſiſtent with ſenſe and analogy as it would be ta write TEYOTE for ETrurk, the T being, in the old Alphabet, a letter as much as the x, ud. no otherwiſe liable to licentious and arbitrary tranſ-

poſition, * » Nez y 8 +.

From the ſecond of theſe two verbs comes ' addon, properly FEFA- - AHKOTEZ, the plural of the participle pana, of FAAELI,\ of. which the

8 %

(1) To ut 764 dad, * ts 78 di. Schol. Ven. ad Il, . 340. See alſo ad it N. 543»



- conſonant has been doubled to ſupply the defect of the aſpirates, as in the adverb addy, which was originally written FAAHN, and the ſubſtantive F AAOE; for the initial aſpirate is ſufficient to extend the firſt ſyllable. Ariſtarchus ſeems to have ſeen the irregularity of doubling the conſonant. and the propriety of adding the aſpirate; but, not being' acquainted with the F, he propoſed to read &9yv, which would fignify at pleaſure inſtead of

to ſatiety; and therefore might do in ſome inſtances,” as uwboreg edpueyar eddy (1), &c. but not in ſuch as adm a xaxoryro;, Kc. It was, how- ever, as before obſerved, a peculiarity of the Molian dialect to drop the aſpirates on ſome occaſions, and pronounce the conſonants double, as in TMMEE and AMMEZ for FYMEIZ and FHMEIZ; and, as this variation 7

muſt have extended to the F as well as the T, Homer might have em- ployed both forms in the verb and adverb as well as in the pronoun. How far he did ſo or not, the analogy of his metre is our only criterion for no light is to be obtained from the antient grammarians. 8 According to Heſychius (2), ſome antient interpreters did not allow ac, in diparog arm wpnu, to belong to this verb, but deduced it from the ſame root as ace, a, «raps, &. I doubt, however, whether a verb ſo derived, and of ſuch a meaning, could, conſiſtently with the Greek

idiom, be uſed with a genitive caſe. Aow or FAZEIN, the future iufini- 

tive of FAAQ, is in other inſtances employed with a dative (3). The adjective aro, properly AFATOE, inſatiable, is derived from FA- AN. It is employed as the characteriſtie epithet of Mars, the god of diſ- cord and deſtruction, whence AFATH, the feminine before treated df, be- came the title of the goddeſs of miſchief, and was e en 3

ſignify m iſebief or misfortune in general,

III. 1. aw audio Ala.

 2. ou * exbals AFIQ. 

IV. 1; aw or aw coacervo FAAQ or FAAEQ,

2. &X6w „ Woe i AAR  

3. N OF a evito AAkra. 4. aAuxu Of ahuoxw Cvito AATKQ or AATZEK . (1) Schol Ven, ad II. E. 203; and Kk. 88. (a) In aK.

(3) See II. I. 81). FLY M 2 5. avoow



3. Hue, Tabie aus ſum, AATEED, in the Attic FAAYEER (1)

6. ave, mens turbatus fum, AATFL, according to others FAATO (2). 

and Y AATIa (3).

7. an, hows aha, cecutire vel errare facio, ANAFQ, AAOFQ,

and AAAFEN. 8. dow capior FAA a. 9. a Or e fun do AAOFAn. 10. 4 ſalio FAAAN,

From the firſt probably came our word WALL, through the medium of the Latin vALLuM. According to the preſent orthography, it is aſpi- rated in ſome tenſes, and thus confounded in the flexions with the tenth, FAAAQ; and, to complete the inconſiſtency, the note of aſpiration is placed upon the augment, ſo that we have iaxg for EFAAH, notwithſtand- ing that the metre requires the aſpirate at the beginning of the ſunple forms, and not of the augmented ; as in

Axa axu; pier (4), properly AXIAEFA FAAEIE MENEN ; :

and Away; 8” da (5), properly AINEIAL A' EFAAH (6).

The ſecond occurs only in the derivatives AAETPIE. and AAETPEFa, which ſufficiently point out the form of it.

From the third are derived aαεν or AAEFOPH, and arm or AAEFH, refuge or evaſion; but whence ane, warmth, is derived, or how it was written, is difficult to gueſs, as it occurs but once in Homer, and there ſeems to want the aſpirate—47v xs Tuzor Jepew, any 16 rat (7). In other writers, however, it is frequently employed; and Heſiod has ra aroxy all which confirm the preſent form. Perhaps, inſtead of digte, we ſhould read OEPEQNM' (that is OEPEQMAT), for the ſenſe ſeems to re- quire the middle or paſſive voice in this verb as well as that which follows.

We have, however, the correſpondent forms of other verbs with the _

vowel equally ſuſtained before another vowel ; as in ov xu 499 Sno ug ay ayw 7 nog Axawy (8),

(1) Euflath. p. 1636. B. a8. (2) Ibid. (3) Ibid. ad Od. I. 398. (4) 11 ©, 591. (s) U. r. 258. (6) In the Heracltan Table, the aſpirate is dropped from da, a congregation z which, according to this hypotheſis, ought to be FAAIA, (7) Od. P. ag. (8) 11, B. 231. 7288 5 1 and


and xypp 3's wah- Ayapipron (1)

no do I recolle& au inſtance in which. the firſt vackon, active occurs in this form, and with this potential or conditional ſenſe, followed by a gonſonaut. Hence I cannot but ſuſpe& ſome corrupt apocope or abbreviation, though I have no more probable emendation to offer than ſubſtituting the common form of the potential mood, which I ſhould not deem admiſſible without authority; for though the change of obfolete to common forms has been general and uniform, it is very improbable that a corrupt change of one common form to another ſhould have. taken place uniformly through fo many paflages, and been ſupported by the concurrent n of ſo many copies and editions.

The fourth is only a different theme of the third, ſuch a as were perpetu-

ally ariſing in the infinite flexibility of the Greek tongue. The fifth is derived from ATE ZA, the canine madueſt, and ſcems to tk a near affinity with the fixth, AATFQ, though differently applied. Ane probably. are ultimately derived from AYFQ. The ſixth ſeems to be derived from the re AAArox, compoſed of the A privative and the verb AAF . 2 2

The reduplication of the firſt ſyllable in 1 the third theme of 4 I have. ventured to conſider as a corruption, introduced, like many others of the

ſame kind, to fill the metre when rendered defective by the loſs of the

aſpirate. In its preſent form, it ſeems to have an affinity with adanyrvg,.

the military ſhout (ſo called from the exclamation AAAAA, whence came

the verb AAAAKNQ), which is, however, a word of a totally different. claſs, being one of the very few employed by Homer not of Greek extrac= tion, and perhaps the only one that can be ſuppoſed, with * degree of

probability, to have come from the Eaſt, The eighth is regular and unvaried through all its flexions ; — the ninth

might be AAOIAQ as well as AAOFAa, did not the ſubſtantive, which is

uniformly «@>wy or orn. and never ag or AAOIH, point out its true form.

The tenth is written, in different tenſes, with and without the aſpirate,

but uever occurs in Homer with the two Lambdas. The occafional omiſ-

(4) 1. I. 387, fion

, P 4 1

  • .

=— :


ſion of the aſpirate is probably a licence of later times, though it may have been dropt, as the conſonants were elided, in particular . or accord-

ing to particular dialects.

. 1. ot ah Tollo AFEIPQ, AFEIPEN,' be. 2. 72 pte, fat, ach, capio, fumo AFEIPEN, AFAIPQ, &c. 3. & convento F APH.

The two "Gel being equally abbreviations of age, antiently written AFEIPQ or AFAIPA, were probably originally of the ſame form; but the aſpirate having been dropt in ſome dialects, aud the initial vowel in others, two verbs were formed, differing a little, but very little, in meaning, for our verb TAKE comprehends every ſignification of both. I ſuſpect, how- ever, that they were not diſcriminated in Homer's time, but that both were written AFEIPQ, &c.; for the vowel is never ſuſtained before the ſecond or aſpirated form; and the contraction in the flexions is perfectly regular.

Upon a very antient medal of Thebes, in the cabinet of the Author, is the word ETF APA (i), the contrafted Aclic or Doric genitive of EYFA- PAE; which, whether it be the name of a magiſtrate, or title of a deity, muſt, I think, be derived from the third verb, the perfect tenſe of which 16 now apypz, and the correſponding participle both apypws and wexpus. Tt appears, however, from the medal, that the original form of the verb was, as I have ſuppoſed, FAPQ, and conſequently its regular flexions FEFAPA and FEFAPQL, the penultimate of which, beginning with an aſpirate and ending in a liquid, may, on either account, be either long or ſhort, and thus ſupply both the metre and ſenſe without any anomaly, We have, however, , with its paſt imperfects ago and apapioxov, and alſo the participle ap#gw ; but, in all inſtances where they occur, FAPQ, EFAPON, FAPIEKON, and FAPAN, will equally fill the places,

The derivatives of the ſecond and third verbs were of courſe aſpirated conſiſtently with the roots; but it is not always eaſy to diſtinguiſh them from each other; for, as the one was uſed metaphorically to ſignify choice, its meaning approached that of the other, which ſignified fitneſs. The

one aſpirate having alſo been totally loſt from the Alphabet, and the other

11) See Dutens, p. 158, where the ſame medal is publiſhed, ſunk


funk into a ſort of accentual mark; applied according to certain whimfieal

rules, independent of etymology, ſuch corruptions have taken place, that 

it is impoſſible in every inſtance to aſcertain the original orthography, more eſpecially where ſo little in formation can be had from the metre. The ori- ginal form of the ſecond, indeed, being AFEIPQ or TAI, naturally produced the derivatives AFOP, AFOPTHP, METEFOPOYX, ' &c. now written aop, aogrne, pernoprg, &c. the A in the two firſt being either long or ſhort, and the E in the third long on account of the aſpirate. Ache is alſo from the ſame root; but «ogu ſhould be from the third verb, though uniformly written with the common aſpirate. Agri, apruvw, &c. are allo from the third, and therefore ſhould be FAPTIA, FApTTNa, &c. unleſs. the aſpirate was habitually dropt in theſe derivations, which was probably the caſe even in the time of Homer, for. I do not recolle& an inſtance where it is required to ſuſtain the metre. Apigov, prandium, was, however, pro- bably FAPIETON, and thus diſtibguiſhed from APILTON, optimum, and not by the firſt ſyllable being long, as Clarke has ſuppoſed; The inſtance of avopigo with the ſecond ſyllable long, cited by him from Ariſtophanes and Theocritus, is wholly irrelevant; new habits aud different dialects having in their times totally changed the pronunciation of the language; ſo that he might as well have cited a word from Pope to prove the right accent of a word in Chaucer. All the manuſcripts and printed editions have uniformly eTuwerrs d , and not evruvorr' apigoy, as he has given it; where fore I conclude that the true reading is ENTYNONTO FAPIETON. VI. 1. on precatis ARE; . 2. Gp") noxa FAPH or TAP.

The ſecond word is always preceeded by a vowel or the paragagic N, ex- cept in one inſtance, and there we ſhould probably read AMTNE FAPHN or FAPHN for apwov «py, the verb being an imperative. Hence we may

ſafely conclude that it was aſpirated ; but whether with the For T is very 

much to be doubted. The Venetian Scholiaſt certainly favours the former when he ſays that wgauog, ſlender or ductile, ought to be written apaucg, otherwiſe it would ſignify. noxious (1). Our word wax alſo, apparently derived from the ſame root, ſeems to ſupport this orthography ; though


0%) 10 U. 2. 411. 5 there


there is no reaſon to believe that Aprg was ſo written in Homer. It muſt, however, be of the fame exttacl jon, and was probably terminated originally in EFZ, whence the oblique caſes are apyog - v, properly APEFOE -EFI-EFA. The accuſative ap, which occurs only before a vowel, is a corruption of APEF* (1).

VII. Ay -nmr0;, and wpyns -rreg, are generally ſuppoſed to be the ſame word ; but, nevertheleſs, I believe they are totally different; for I know of no licence that can double or extend the penultimate vowel in the oblique caſes, without the aid of a liquid or aſpirate, but what would ſubvert the analogy of all language. They are applied too to objects ſo different in their natures, that it is ſcarcely poſſible that they ſhould ſigniſy the ſame properties; for when the penultimate vowel in the oblique caſes is long,

the epithet is always joined to ſomething ſplendid or agreeable, as wpyyra

up, lava apyyri pany, &c.; but, when it is ſhort, it is never employed except to deſcribe the fat of a dead carcaſe, as ace ev Tpoiy T&x,60G Kuvas age ver ng. It is poſſible that the firſt might have been written APTE FS EFox, and ſignified emitting whiteneſs or ſplendor; for moſt, if not all, words ending in EFZ, were ſignificant of action. The other might have been written APPEE -ETOYL, and have ſignified the dead inactive quality of whiteneſs, This diſtinction may perhaps appear refined; but ſuch re- finement belonged to the Greek language: thus EINMEIOL or EINNIKOE ſiguified any perſon or thing which paſſroely belonged to horſes ; but FIMNEFS the perſon atively belonging to them, that is, he 2vho rides or drives them.

VIII. 1. daw accendo AAFQ.

2. J divido Ala.

3. Jaw dlſco AAkEn. From the firſt come AAFIL, AAFOL, AAFIOL or AEFIOE (employed metaphorically to ſignify defiruttive in general) ; and thence AEFQ, AE- FION, &c. which are all written in the antient manner, except that the

aſpirate has been dropt, and the H introduced inſtead of the E, to ſupply

its office in giving the ſyllable its. due length. The 1 has alſo, in the flexions of the verbs, been joined, in the form of an Iota ſubſeriptum, to the preceeding inſtead of the ſucceeding vowel ; whence we have Jywoas,

(t) See II. z. 100. &c.


ccc. inſtead of AEFIO TAZ, Sc. Aye I believe to be a corruption of AE» - FION, formed according the common mode of the Attic contractions; but nevertheleſs, AEFIOON, the regular flexion of the more uſual theme, may equally be a word of two ſyllables, and therefore the true antient form. The firſt ſyllable of the adjective Jyio;"is frequently ſhort, which proves in- diſputably that ĩt was antiently written AEFIOL, the firſt ſyllable of which might naturally be either long 'or ſhort, whereas no licence could ſhorten a double vowel in this place. Ae and eg are from the ſame root, and were originally written AAFEAOE,. both being the ſame word, employed literally and metaphorically, and contracted, according to. different local idioms, to AAF AO, AAEAOE,. AEEAOZE, and AHAOE, Hence we find in Heſychius AABEAOYL, dag, AA,, AAEAON, Na, and AEEAON, . Aai is ſuppoſed by Damm and others to be a contraction of Jai, or AAFI- Al, the dative fingular of dai or AAFIE; @ torch ; but as it is always uſed, when thus abbreviated, in a ſenſe which it never ſignifies when at length, I ſuſpe& it to be a different word from the ſame root, the regular flexions of which would be AAFE -AFOE, AF, &c. ; though, as it only occurs in one caſe, the analyſis-of it cannot be very certain. It is evidently ein- ployed metaphorically to ſignify a fight, in the ſame manner as AEFIOE or AAFIOE (which ſeems to be the adjective e formed from it), ls to ſignify deflrufiive. 1 From the ſecond verb came AAIE -ITOE; a > ſeal 4 or entertuigment, ſo called becauſe the proviſions were: always divided regularly among the gueſts, Hence come various other words, ſuch as" AAINYMI,, AAITEO= Mal, AAITPOZE, &c. which: will be all found conformable to their roots through all their flexions and variations. : The third is regular through all its flexions in x the preſent orthography

IX. 1. % Iigo -_ 1 0s - „ W 1 nent a; nk 5 1 3. dew. , egev.. EF. 4. Mu rige AETQ.

nn 5 AYFQ.,

The Lorna and flexions of theſe verbs are obviouſly poigited out by — fect and metre. From the firſt came AHMOE or AEEM Ox, à people ; and from the fourth, probably, &quos or AETM OZ, fat, which ſome an-

vb N ids hep tient 


tient grammarians, however, derived from Jaw or AAFQ, to burn (0 in which caſe it muſt bave been written AEFMOZE; py rag

be right.

Ni * 1. * Firgio ASI or Aan. "= F 2. dw or bw fugo Ala or AlEa.

+ The forms of theſe two verbs are diſtiuguiſhed by the angle vowels-

preceeding the former being always long, aud thoſe preceeding the latter

always ſhort; otherwiſe they might be really the GG n.

times in a neutral and ſometimes in an active ſenſe.

The firſt has been already very fully examined, and the ſecond has no- thing particular in any of its flexions. It feems to have been doubted: among the antient Criticks to which the word employed by Hector in- II. k. 251 belonged; whence ſome editions gave it in the firſt perſon, Jv, and others in the fecond, Jug, the latter of which was AY

n though the ſormer prevails in our modern copies.

XI. 1. . capio FAN. 2. dw Or ww volvo, congreco FEAQ_Oor FEIAQL The flexions and derivations of theſe two verbs have been much con»

founded by the modern reſtorers of the Digamma, who, becaufe both be; gan with a letter capable of ſuſtaining the preceeding vowel, concluded! that both began with that aſpirate (2). The one, however, being already aſpirated, there is no reaſon for altering it, eſpecially as the afpirate ſerves to diſtinguiſh it from the other, which is ſo different in meaning. The fame may be ſaid of its derivatives FEAQP, BEAAOMAI, &e.; from the latter of which, indeed, the afpirate has been dropt in compliance to the abſurd rules of the ſchool-reformers of the orthography, though- both are equally derived from EA (3 and ſhould * be written ac- cording to etymology. .. -

FEAQ or FEIAQ, having begun hk a 10505 ug obliterated and for- gotten, has of courſe been more diſguiſed, though not ſo much ſo as to be very difficult to be traced iu any of its flexions and derivations. Dawes has

remarked that «aca ſhould be FEAEAI, and eamungy, FEFEAMENOE.; and 

conſequently the ſame analogy muſt regulate the orthography of ry

(0) Schol, Ven. in II. ©, 246. (a) bes Dames ds Cn, vl Apr, Va. (3) See Euſtath. P · 36. J. 42.

1 g word

tenſes, EFEAIKZE, EFEAIXOHEAN, &c, Hence the laſt ſyllable of vu veog is extended before , not by any ftreſs or emphaſis upon the &, but by the natural effe& of the F.—KYFANEOE FEFEAIK TO APAKAN(I), The imperfet might indeed ſeem more proper in this place ; but the paſt perfect, OPAPEXATO, having been employed a few lines before to.cxprofs the ſame time (a), proves that this muſt be equally a paſt perfet; EAF-» Alara, to ſbate or vibrate, is a word of a different extraQtion, though confounded in the flexions with 1 10 curn, by the defects 4 11

Foe orthography. Py From EAN and FEA are derived o&o3g or whos, properly +OAOFOE or

10Aror, baneful or deftrutirve, and v, afterwards de, properly ro-

AOFOE or FOAFOYZ, cullected or whoſe, aud metaphorically curled or ' avoolly, whence the Latin verb voLvo was formed without any change but

that of the v to its correſpondent letter in that alphabet, When the firſt

adjective is in three ſyllables the penultimate is in a very few inſtauces long, whence the Venetian: Manuſcript has it with the diphthong, as ιu od ble emiduew (3) 3; and 1 yap dy ohni gps. Jun (4); which may poſſibly be right, though OAOFH is the more regular form, the penultimate of which may be either long or ſhort. It ſhould, I believe, be pronounced

in the ſame manner in ſome inſtances where the contracted form is now.

employed, as in du, lu yag py , aryp—— (5) i Agapapmors U og (6); and facx' d. u. overpe (7) ; which ſhould probably be AEAAIA TAP H- FOAOFOE ANHP——; ATAMEMN' FOAOFON ONEIPON=— 3 and BAZK”

10 FOAOFOZ, ONEIPE., In this laſt inſtance I would ſubſtitute the no-

minative on account of the metre, as in g w MI -; a Qinogy Yave xo ou, &c.; in which eg is not an Attic vocative, as ſome have

ſuppoſed, nor a nominative put for a vocative, but a nominative regularly joined to a vocative by means of a verb or participle underſtood, the ex- preffion being elliptic for 5 u or p wep t: In the fame manner

BAZK' e' FOAOFOZ, ONEIPE, means literally go baneful, O dream! or

' (1) 11 4. 39. (a) Ibid. 26. - (3) U. x. g. (4) II. A. 342. (5) II. 6. 536. :


ON THE GREEK ALPHABET. „ word derived from it. Ex# ſhould be FEALK Z; dvi FEAIEER 3 aud the reduplications AAge, NM &c. only the regularly augmented


, collect ively on bodies or troops, wiape;, a mob or crowd, unf, a fur-


go dream that art baneſul (1). It is poſſible that the initial aſpirate might” have been dropt from this adjective even in the time of Homer, and that it might have been written OAO FOR and OAFOE; from which the verbs Ora, OAFTMI,' and OAFEk Ka, are derived, as vol vo is from FOAFOE, and of courſe ſhould follow the orthography of the root, except that the F is elided in the flexions as the ſecond T is in the flexions of Trura. EXvog, a table 10 turn or dreſs meat upon, and «og, mercy, were antiently diſtinguiſhed in the fare manner, the firſt being from FEAn, and written FEAEOF, and the ſecond, a word of leſs certain etymology, EAEOSL. "Emeg, a bog, ug, mud or clay, ue, (or, as in the Venetian Manuſcript,

row, the name of the city Ig, &c, are likewiſe from FEAQ or FEIAQ, and ſhould conſequently be written FEAOZ, FIAT E, FIAAAON or FEIAA- AON, FOAAMOE or FOAFAMOZ, FOAKE, FIAIOZ, &c. by which means

the metre will be rendered correct as well as the etymology diſtin,

Ach and ae ſhould alſo be FAFOAFOZE and FAFOAFIAEN, forthe

_ A, prefixed, being what is called the afpoigmeov, or collective, ſhould be aſpi-

rated (2) ; whence Ariſtarchus aſpirated abpoog (3), as the Attics did ab pee (4), and others atr\pog, aaf, abun, and all the words of this kind (5) from which the initial letter had been dropped through local and habitual corruption, and the defects of a new Alphabet. It was dropt from FA@POFOE for no better reaſon than beeauſe an afpirated conſonant fol- lowed, which was contrary to the rules of the later grammarians (6), For reaſons equally frivolous it was probably dropt from «dwg, which Damm would derive from the ſame root as au/; but the ſenſe in which it is al-

ways employed ſhews that the initial is the A abpoiger, and therefore, that

it ought to be preceeded by the aſpirate. Similar corruptions ſeem in ſome inſtances to have taken place in affixing it ; whence probably the A priva- tive 18 atk HY in iKatro, —_ to general . and, apparently;

i © Dames ſuppoſes that v in both theſe juſlances means whele or entire ; but 1 think Clarke's interpretation, which is alſo that of the Scholiaſt, better.

(2) Euſtath. P. 16, J. 3% (3) Ibid. P+ 996, I. 10. (4) Ibid. 1387, I. 7.

(5) In the firſt Sigdan Inſcription we find HASEASOI.,

-(6) Euſtath, p. 1387, 1. 1. Þ


to the practice of Homer (1) ; for in the numerous paſſages where this verb is employed the aſpirate is never required by the metre. It appears alſo. from an obſolete word, which occurs only in ſome copies (2), and which has been explained by the antient Critics, that there muſt have been, two verbs of this form, very different in meaning, and only diſcriminated by the aſpirate, IAM APT from FAMA and APTAQ or FA Ta; and AMAp- Ta from A and MAPH Ta. XIL 1. eg, Ep8cw ON eig dico 'FEPQ, FEPEQ or FEIPEQ. 2. 8pwy p Or eptrw rage -EPQ, EPE or EIPEQ.

The metre points out the different forms of theſe two verbs, the firſt requiring the aſpirate to ſuſtain the preceeding vowel in almoſt every in- ſtance, which the ſecond never does. The derivatives, however, from the firſt are proc, pp, pyrwp, &c. but as the Laconians wrote them with the B, BpHT OZ, BPH THP, BPH Ta, &c. we may conclude that the old Ho- meric form was with the F, FPYTOE, FPHTHP, FPHTQP, , &c. the re-

gular contractions of FEPHTOZE, FEPHTHP, FEPHT QP, &c. which were

the regular nouns, formed according to the common rules of analogy from the verb. In their preſent forms they appear to be from pw, uo, which was antiently written PEFQ, whence came POFOF, contracted to POP, and now written gu. The antient grammarians and ſcholiaſts found a dif- ficulty in the flexion of the word Xupages; ; for not all the licence of con- traction and extenſion, in which they ſo freely indulged themſelves, could deduce from it the plural x Hf (3). The caſe is, that this word has ſuffered a double corruption; firſt, by omitting the ſecond O and ſubſti- tuting the T for the F, and then by doubling the P to make the ſecond ſyllable long. The true word, formed according to the regular analogy of the language from XEIMA and PEF , is XEIMAPOFOE, the regular plu- ral of which is XEIMAPOFOI ; and I believe that, if theſe: forms be adopted inſtead of the preſent, in every inſtance where Homer uſes it, the deſcriptive beauty and rapidity of his metre will be as ben improved as

(1) Heſychius has — as before obſerved, an A privative with . F; but this was probably another local corruption. (2) See MS. and Schol. Ven. II. E. 656 ; 2. 571; 0. 162 and Ye 414.

(3) See Schol. Ven. in II. 4. 452; 4 the


the regularity and preciſion of his grammar, If it was e e it muſt have been to XEIMAPFOE.


Eee, co in perniciem, is of an extraction different from any of theſe, though it appears from Heſychius to have been written FEPPQ (1).

XIII. 1. worw Jum Ea or a.

2. 4% [ew Or Paw, vado FEN, BEQ, or BAN, the Laco- 

nian idiom having in this inſtance _—_—_ * mitto EN. deo, vel

1 we — ua. uſtulg FEFQ, veſtio Fa.

IT To theſe ſome add ; iu, expleo, and thence derive ic Or 5ojuey (z) Eck

occurs only once), and ire (3), but the laſt is from H, the form in -MI of Hin, and the former is probably a corruption. The Scholiaſt ſays, that ſome antient editors gave xrouw{4), fignifying Snoxwproouer; and this is probably right, x or XEFOMEN being the Ionic mode of writing. XAFOMEN, from XAFQ, the primitive form of yaZu, whence comes u or XEFA, the retreat or hole of a ſerpent. X. or XEFQ is uſually employed metaphorically to ſignify pouring out, whilſt XAFQ and XAALQ retain their primitive ſignification of yielding place or vacuity, whence came XAFOE, XAEMA, &c. In Od. Z. VI. 17, we have how- ever, olog I" ahh eg ad yeuriray this threſpold will afford room for us both. The verbs which I have collected under this head being compoſed en- tirely of vowels and aſpirates, the moſt flexible and variable parts of the moſt flexible and variable language ever ſpoken by man, they have naturally varied their forms more than any others, fo that it is exttemely difficult ts trace every flexion to its proper theme, and ſtill more ſo to diſcriminate the corruptions. of later times from the cuſtomary dialects of the Poet's own age and country. Each of them has its termination in -MI, as indeed

' (+) In BAPPEI and TEPPNA, that is, FEPPN. (2) Exe x ide wamoios II. T. 402, al. ihn (3) EE ven ivroy ' expelled the defire. N (4) Ene xt mohyacucs every


every other verb had; but in theſe it was ſo prevulent that the” we, n form grew obſolete, except in the flexions, at a very early period. The forms of the tenſes in the firſt are ſtill quite regular, except 4 the future «cope is from the paſſive, though the Aoriſt ' yew, we, or yo, it from the active. The imperfect ww is uſually replaced! by e, from the termination in - which was alſo pronounced , the middle vowel. be- ing made to coaleſce indifferently with the ſucceeding one; or with the pre - ceeding augment. When the augment was omitted it was u, as the Ao» riſt was «oa, which. being the Atric forms, became general for the im- perfect, the Aoriſt having grown obſolete, except in the third perſon phs- ral, «ow, and that was adopted for the imperfect. Hy, which occurs only ence in the Hiad (v), and twice in the genuine parts of the Odyſſey (2), is the paſt perfect, regularly n. Other variations, ſuch as the E in the fu- ture and Aoriſt being pronounced double, fingle, or not at all; the ſecond perſon 6ygular being eontracted or at length, &c. will point out themſelves to all who are converſant in the language, aud wilt confider it according to its analogy, without placing too much confidence in our common ſchaot- grammars, where he may find the antient regular Wan of the tenſes given as licentious deviations of diale&, . Taye ſecond varied its form to EFI, ra, &e. from which various flexions rn were formed, The Latin vivo, however, which is the ſame word in a different diale&, proves that the original was written with the F. Our old verb wzNnD too, of which we fill uſe the paſt tenſe wzwr, ſeems to have come from FEN TI, the antient form of the third perſon plu- val of FHMI. Hence likewiſe came FE TO (now. written «ro, but in the Heraclean table EETOZ, and in Heſychius FE TOF), 4 year, or period of the going round of the ſun (3) 5 and FEANON or EFEANON, @ marriage portion, probably formed from rhe antient theme FEAQ, correſponding wirh the Latin vano, whence comes our word web. The adfcititious E, n ohne . both in verbs and ſubſtantives, is not, as Lennep ſup-

(1) A. 80%.

(a) T. 283, and g. 342 3 the recapitulation in v. 410, &a. ſoems to be ſpurious. 5 (3) Hence, as Mazzochi has obſerved, came the Latin word vErusr us, of which vzrus- a contraction; the old Greek. being EETOZTOx, literally .


1 ..


poſes, an arbitrary prefix (for no ſuch Bees can exiſt in any language); but marks a new theme, formed from an augmented tenſe, ſuch as the Greeks were continually in the habit of making. Erie, 4 fellow-citizen, ſeems, in its preſent form, to be of the ſame extraction as ereg or FETOx, though it is in reality very different. Some antient grammarians ſuppoſed it to be an abbreviation of i rapoc, and that therefore it ought to be written

  1. 715, unleſs the aſpirate was dropt by local or temporary babit (1). It ap-

pears, however, to be rather the root than the derivatiou of trages; and that it ought to be written with the aſpirate, we have the undoubted au- thority of the very antient Leſbian medal, on which we find the genitive plural FETAION, either as the title of the local deities repreſented upon the coin, or as an appellation of the citizens, in the ſame manner as IlOA- KOZE (a word of the ſame origin as the Latin voLevs, and our FOLK). is employed for An Mo upon a medal of Cnoſſus, in Crete, belonging to the collection of Mr. Cratcherode. From #ry;, or at leaſt from the ſame root, comes ere, true or certain, which therefore ought to be FETEOYE ; but race, vain or tranſitory, muſt be from FEQ, and therefore written with the other aſpirate, FETQZIOEZ. From the'form 1Q probably came the adjective 5 lou, regularly FOMOITFOL, that which comes equally upon all, which 1s therefore pronouticed in four ſyllables with the penultimate long ; but % arid u are from EFIQ, and were therefore HFIQAN and HFIA; whence the laſt is ſometimes pronounced in two, and ſometimes in three, long ſyllables, HF-IA aud H-FI-A.

The third of theſe verbs exiſts ouly in the termination in 1 MI, FHMI and HEMI or EIEEMI, though it ſeems antiently to have been in -K, FEKQ, whence comes the Folic Aoriſt zyza, properly EFEKKA for EFEKEA, Ac- cording to modern orthography, the ſimple aſpirate, now ſignified by the mark (“). ſenroely ever occuts in the middle of a word, except witli the aſpirated liquid p, which being neceſſarily pronounced with a forced as well as conſtrained expiration, does not want it, and therefore never has it in any ancient inſcription. On the pillars of Herodes Atticus, however, in- ſcribed under the Antonines in imitation of the very antient orthography, we have ENFOATA for woda, and, in the Heraclean inſcriptions ANFE-

(1) See Schol. Ven. in II. Z. 239, 7 f axel,


axel, NAPHESETAI, &c. for ane, waptira, &c, all written ac-

cording to etymology ; which ought therefore to be our guide in this as well as other reſpets. Awprs ſhould be AEQPOS, ab,, AFTIINOL, &c. &c. Even in the flexions of the verbs, when the T was elided from the

ſecond perſon ſingular, the ſoft vowel aſpirate F was ſubſtituted in its

place, as NOIRFAT for NOIHEAT, now written w, (1), which may ac- count, in many inſtances, for the metrical quantities being ſuſtained, 1 have hence veutured to ſuppoſe, that the verb in queſtion ſhould be writ- ten with two aſpirates, by which it is-not only diſtinguiſhed from others

of ſimilar form and different meaning: but a reaſon is given why the firſt

ſyllable of the participle ih is uniformly long, it having been written antiently HEEMENOZ. From this form probably came n, which, in that caſe, muſt have been regularly FIFOFH. lov, or (as in ſome editions) zor, a miſſile dart or arrow, is derived from this verb, and therefore ſhould

be HEON ; as wy, 4 Violet, ſhould be FION, according to the Latin vioLa,

derived from the ſame root.

The fourth verb of this head is varied to an and Hara or klxaa, the regular flexions of which frequently occur. The fifth I have ſuppoſed, from general analogy, to have been written Kyra rather than RT; but the preſent me may nevertheleſs be right in this inſtance,

The fixth, I venture to conclude, was written with the F rather han the E, becauſe the Latin word vssT1s is evidently Yerived from it, and

becaufe the Laconians wrote it with the B (2); but, as theſe ſimple aſpi- rates were fluctuating, and varied with local habit, it is impoſſible to de- cide, without better authority, in what manner Homer wrote it. Eaveg,

a robe, is certainly derived from it, as ive, ſubtile, probabl y is from ta or EFAQ, 0 leave or permit. At preſent theſe two words differ only in the quantity of the ſecond ſyllable ; a difference for which there is no ap-

written FEANOZE, and the ſecond EFANOxE, by which the form becomes

  1. 8 different as the meaning, and a ſufficient reaſon is given for the A being

invariably ſhort in the one, and invariably long in the other, . ves

(1) Etymol. magn. in Voce io Viltoiſon, Proleg. in Homer. «>> 2. (2) n magn. & Heſych. 0


parent reaſon ; but, if my conjecture is well-founded, the firſt ought to be


&c. are from this root, and therefore ſhould be written, as before obſerved, FEOOZX, FHOOE, &c;, Hence BELTON or BETTON - ſignified, in the Laconian diale&, both a manner or cuflom, and a garment, by a power fimi- lar to that which the word #4aBrr has in our on language (1). XIV. 1. %%%, perf. ih, opere, FEPF n, perf. FEFOPPA. _ 2. t, perf. uyyz Or tig, arceo, includo, EPF, perf, RR. TA or FEFEPXA, as in the Heraclean Inſcription. From the firſt come FEPFON, AFEPLOE, &c.; and from the ſecond, FEPKOZ, FEPMA, &c. which are ſtill written with the afpirate; though it has been dropt from their root. Dawes perceived that an afpirate or cconſonant was wanting to each, in order to ſupport the metre ; but, not attending to the metrical power of the E, nor taking the trouble to trace the different derivations, he prefixed the F to both, and thus confounded two words as different in form as in meaning. The ſecond was often written with what Lennep calls the adſcititious E (which he ſuppoſed to have been arbitrarily prefixed), EFEPIQ ; but this is a new form from the augmented tenſe. It ſeems alfo to have been written with the old Ionic, or, as the grammarians called it, the Attic I, IPT a, for in this dialect they allowed it to have been aſpirated (2). They have not, however, in- ſerted the aſpirate in -@T047 (II. ©. 283), though. the natural means by which the ſecond vowel was made long, AMOFEPEEI ; for, that ſo it ought to be written, and not with the F, as Clarke has conjectured, 1s proved by the firſt word of the preceeding line:

b Texlur e ννẽᷣ Her,, we Saks ouPogoy 

"Oy 2 r Wouhes aroupor x oppor a. XV. 1.4 1% an? vel ET Erk or H, HE.

5” WA W.4 O 2. 7 certè Hor EE. ht Cn & - The firſt of theſe words, when in one ſyllable, is ufually ſuſtained before FO 45 „ A vowel; whence Dawes concluded that it was written EF. In ſome in- 1 ſtances, however, it forms but one ſyllable with the negative vx ; whence

I am rather inclined to believe that the regular form was always in two ſyllables, H E, and that the laſt vowel is elided, in the caſs eee o as to form -

(1) Etymol. magn. & Heſych. nnn Mevrſins in Laconie. lib. bg c. vis. - (2 See Schol, Ven, B. II. o. 282, '



XVI. 1. den or % Ne ?&dꝛła or ea. 135 2. 9% curro EFW. 9 3. dee or Jarw Video, miror oOkRq or ARG.

Wa ths; augmented tenſes of the firſt come 761u; and. Inzw or OEK , which more frequently occur: but dere, the third perſon ſingular. of the imperfect or ſecond Aoriſt middle, can only be from the original theme; from which, as Herodotus obſerved, is derived. the word . the firſt ſyllable of which is therefore invariably ſhort, I have ventured to conclude that the ſecond was written with the aſpi pi-

rate, not only becauſe it would naturally be diſtiuguiſhed from the firſt in a primitive language, remarkable for its perſpicuity, but alſo becauſe the two is in Yes, the third: perſon ſingular of the imperfe& or ſecond Aoriſt,

never coaleſce into one ſyllable ; and the ſecond perſon of the future mid»:

dle is Neues, and the infinitive Sue, which, I think, can only be corruptions of GEFEEAT, eEFTE TAI or Ag. and eEFPZExeAI, by the uſual change of the F into an 7.

The third has been much diſguiſed by the ntrodudicn of the 3 VOW= els, and omiſſion of the aſpirate; but, nevertheleſs, the original form is diſcoverable, by regularly 4racing the analogy in every flexion; aud may

in every inſtance be reſtored, without violence either to the ſenſe or metre.

XVII. 1. #6 or x8: bono, jaceo KE or KEIQ. 2. Ne, Kai Or xt ro - KAFQ, KEFQ, or KHFQ. 3. xt, OT Kh Ando KEAF or KEAA Ta.

The forms of the firſt appear to be regularly preſerved ; but the active form occurs only in a neutral ſenſe, unleſs it be in Odyſſ. Z. 425, xoþs

  • aa NI Jouve; q Au N, which Clarke, Damm, and others,

explain, that he-firuck the victim with the ſplinter of oak that he had left of { 6 "a

  • . —

cleaving but, befides that & no where elſe means to cleave, the tenſe ' in which it is here employed decidedly precludes that fignification ; for, according to Homer's idiom, it muſt neceſſarily have been the perfect or Aoriſt, inſtead of the preſent. I therefore believe that we ſhould apply the latter part of the verſe to the latter part of the action, aud underſtand the

blow to have been firuck with a ſplinter, which he then left, laying it down. 

The very learned Chriſtian Tobias Damm held that the neutral ſenſe, in -

the anticat Greek verbs, was the ſame as the active, except that the pro- O 2 noun

{ % DS


noun or ſubſtantive was underſtood, which may account for: almoſt every vorb having a neutral as well as an active or paſſive ſenſe. This is often expreſſed by what is called the middle voice, of which more will be ſaid hereaftes, It is poſſible, however, that the paſſage in queſtion may be corrupt, and that, inſtead of u, we ſhould read KEAAQN or KEAZAE. The ſecond, I believe, ſhould be always written with the F, inſtead of the 1, to diſtinguiſh it from the firſt, whence, in Pindar, it is xauw or k Ara (i). The future, pronounced in the Ionic manner KEFEQ, may,

indeed, appear to be thus confounded with the future of KE Fon, 10 hide, 

which is. now alſo zwow or KEF Za, but is regularly KEFOZNQ. The 2,

however, in the flexions of KAFQ is clided ; whence we have eye" for EKAFZA or EKEFZA; and it is poſſible that this refinement might have taken place even in the time of Homer, for the Ionian Greeks ſhewed their abhorrence of this letter at a very early period. The © in KEF@EN, O@-» ra, NEIOEA, &c. might have been dropt, for the ſame reaſons, at the ſame early period: for we have ſo few monuments of very early ortho- graphy, that it is r l to trace accurately the nne of theſe r re- finements, ; XVIII. 1. N Fange KAAQ. . 24. Aae, Xhavee, Of xa HJamentoy KAAFQ or KAAIN. The firſt-ſyllable of the ſecond is always long in the flexions, and that of the firſt ſhort, which points out the antient difference in the ortho- graphy. 18 XIX. 1. c, gen- Aus leo Arz, gen. AIFOE. 24. Ne, gen. Ne Javis ANA, gen. AINTOE. The firſt occurs only in the nominative and aceuſative ſingular. in Homer, the latter of which is a7«(2) in our preſent copies; whereas it ought to be AIFA according to the rule of flexion here ſtated, In a paſſage of Callimachus, however, eited in the Venetian Scholia, we have the da- tive plural 27400: (3), that is, AIFEZT,. which proves that the N, in the accuſative, is a corruption, intraduced to ſuſtain the 2 755 rendered. de- enn b


(3) Ibid: TI have

(3) See New, X. 65; a) ey Py


I have ventured to ſuppoſe that the N ought to be added in the ſecond, not only becauſe it is a word of the ſame fignification and etymology as

AINON, but becauſe this letter has been dropped, as before obſerved, out of many words, which in antient Inſcriptions-are formed with it, XX. 1. ow fers Ola.

2. ow puls OlF a.

with the aſpirate FOILONTI, for ec

The 1 in the ſecond, being uniformly long in the bang, muſt hare

been followed by the aſpirate.

XXI. I. m F O POF. 
2. % conciio OPA. 

The firſt of theſe verbs is always employed in a neutral 4 0 * the ſecond always in an active one, except when it occurs with the adſcititious augment, as in opcpe: and gpwpe, which are always neutral, and uſually ſig» nify the imperfect tenſe. Theſe ſingularities are extremely ſufpicious, and

nduce me to believe that theſe forms are corruptions of the regular flexions-

of the firſt verb; and that, inſtead of opwps, and opwpe, we ought to read OPOFET and OPOFE. When they ſignify the perfect tenſes they were probably written OPOPFEI and OPOPFTE, the regular augmented forms contracted. - From theſe verbs are derived ſeveral words of ſignifications apparently remote from each other, but which have nevertheleſs a very eaſy and natural 'counexion. Oupog (maſculine), the original form of which appears to have been OPFOL, fignifies an impellur, or exciter 16-aBtlon, and thence a /cader, director, or guardian, in general. Hence alſo. # favour»

which a ſhip is launched or impellud into the water; Thence it was em- ployed to ſignify the ditch or mound that divided the lands of different pro- prietors, and, by degrees, a mounding or termination in general ; to diſtin- guiſh which from its other ſignifications, later writers, and the euſtom of other dialects, ehanged the aſpirate, and for OPFOE wrote FOPOE z but in the Heraclèan Inſcription it is without any aſpirate. Oupoy or OPFON

Ggnifies the a? of impelling or exciting; and upog, OPFOL, or OPOE (neu-

ter), a mountain, that is, a'maſs of earth, which ſeemed to have been raiſed

6 „

The firſt occurs only in the future, which is therefore aſvally treated « 1 an irregular flexion of ERH. In the Heraelèan Inſcription it is written

able wind that impels a thiy; and likewiſe a Jip or channel. in the ſhore, by

  • =

ad ms. _— TT r 9

- * .





or excited from the rei. OPOE is the whey or ſerum of mill, which is pro- duced, or ſeparated, from the coagulum by an add ion fignified by the verb OP. From each of theſe many other words are derived by a regular and uni- form proceſs ; for, as the Greek is an original tongue, the complication and developement of its elements correſpond exactly with the complication and developement of the ideas which they repreſent ; ſo that the ſtudy of it leads to an examination of the firſt principles of the mind, and ſoars above the humble ſcience of common grammar.

XXII, . guw or puew flu, PTQ or PYEQ: 2. gow _ PYF. 3. api trabo, det ineo FEPTa. 4. #pur reirabo EPTKq.

The firſt of theſe verbs ſeems to be only a variation of * from EFA, which was before conſidered,

The three laſt, being ſamewhat ſimilar in their. meanings as well as forms, have been more confounded by the change of the orthography than any others. To draw, withdraw, or withhold, may ſignify almoſt the ſame action; and, when applied to danger, may alſo ſignify to ſave or de- fend. The ſecond is, however, uſually diſtinguiſhable from the third, even in the augmented tenſes, by the penultimate ſyllable being long; for in the third it is naturally ſhort, Whenever, too, the unaugmented forms of the third are preceeded by a vowel, that vowel is ſuſtained $ which proves that it was written with the aſpirate.

The cuſtom, however, of forming new themes from the augmented tenſes has created ſuch confuſion in the flexions of theſe verbs, originally ſimilar in meaning, and rendered ſimilar in form by the loſs of the diſeri - minating letter, that it will be found extremely difficult to. retrieve the antient orthography, in all inſtances, without. the aid of better manu- ſcripts to aſcertain the true readings.

The derivatives gg, gurne, and purog, ſeem to be derived * pack 3 though the firſt ſyllables of thoſe derived from the third being contrac- tions, as in FPHTOE; FPHTHP, and other words from FEPQ ; and the F being loſt both from theſe and the others, the forms have been utterly confounded, at the ſame time that the ſenſe of the context has preſerved the different meanings. Puges ſhould bo always written Frrox, being




derived from FEPYQ, as is gr in H. H. 475, and Odyſſ. Z. 261, and o. 273, Where it ſhould of courſe be written FT TH; but, in Odyſſ. E. 187 and 223, it is evidently from PYFQ, and therefore ſhould be written PYFTHP-- Purog,, which only occurs as an epithet to ſtone employed in building, is uniformly explained by ſcholiaſts and lexicographers to ſignify aduectitious, or drawn from the quarry; and, unleſs we had the evidence of more certain analogy, or antient monuments, to contradi them, the ſaſeſt way is to ſuppoſe that their traditional Information. is right; and conſequently that this word was written FPYTOZ. Eppa, which occurs only once in Homer, and once in Heſiod, is there uſed to. fignify defence, though the metre ſeems, in both inſtances, to require that it ſhould begin with the aſpirate—yy cp spupuc xpoog—and . Tore £00021 tpupor ypoo KAG4vey per pehaxy, &c. Ariſtophanes and Zenodotus ſaw that this was corrupt, and therefore propoſed to read sauue, that is, FEAYMA, from FEATQ, to envelope; but, beſides that no ſuch. ſubſtantive occurs any where elſe, the ſucceeding verb eyvro or EPYETO, in the paſſage of Ho- mer, proves that it muſt be derived from PYFQ. I would therefore ſub-

[Nitunte PYFMA, now written pipe in both paſſages—FHN EOOPEI PYFMA.


-XAAINAN. MEN MAAAXHN, &C. as o@aA4gev- Wveyu eta, in the Ajax of Sophocles, and goua pon Jogos, in Lycophron. E eu is, however; con-

ſtantly uſed by later writers to ſignify defence, they following the text af

Homer as it then was, or, perhaps, adhering to an habitual corruption,

which was ſanctiſied by uſe, though originating in a literal error; for both

the forms of the Digamma, L and F, differing from that of the E only in

a ſmall tranſverſe line, they were often miſtaken for it by the antient tran- ſoribers, even in engraving the moſt ſolemn public Acts, ſuch as the He-

Taclean tablets, which muſt have been much more carefully. and delibe-

rately executed than books copied by trading ſcribes and rhapſodiſts. In the ſame manner, therefore, as. in the inſtance before. obſerved, LETOE

became EE TOR, in an age and country where the L was regularly in uſe,

"PYFMA became firſt PYEMA, and then (to avoid a monſtrous and unintel- ligible word) EPYMA, in ages and countries where it was wholly diſuſed and forgotten, and, therefore, more liable to be miſtaken, In another pallage of. the Ajax of Sophooles, we 18 however, epupece, to ſignify de-

. fence 3.


fence (1); but there, I believe, it was written pps by the poet——rrper ches Tpwwy not being ſo regular either in metre or dialect as woes pipe rp, In the preſent text of Homer we have alſo «vomroay but, as the vowel is

always ſuſtained before it, we 4 7 5 nn that it was originally PTFIIN- TOAIL.

XXIII. 1, pow uo porn.

2. , et FEPOFEQ.. 

Mr. Duwe by a ſtrange inadverteney for à perſon of his learning, would prefix the F to the firſt, and for eggwrayro write EFEPPAZANTO, notwithſtanding that the imperfe& tenſe oceurs frequently in the ſame voice, and at the beginning of a line, without any augment, pworro or PO- FONTO. Epuy, impetus, being from the augmented form of this verb, was written EPOFH, in the ſame manner as the ſubſtantive regularly formed from EPOFEN ; whence it is one of the very few words in Homer

which have two different, and almoſt oppoſite, meanings.

XXIV. 1. ooog ſaluus TAoOx. 2. ooog agitans TOFOE.

The firſt of theſe two adjeRives is derived from ZAC, to ſave, and the ſecond from TOF, 19 ſhake or move violently; and it appears, from the Venetian Scholia, that ſome of the antient Criticks would have diſcrimi- nated them as they are here diſcriminated. The ſecond occurs only in the compound AAFOELOFOL, agitator populi, the epithet applied to the god- deſs of diſcord, and other warlike and deſtructive deities. Neither is the verb from which it is derived ever uſed by Homer, at leaſt according to the preſent orthography of his works ; but both Herodotus and Sophocles em- ploy it in oblique and diſguiſed fortns. Tas (2) and vun (4) Valkenaer would, indeed, make owe:, an abbreviation of o but improperly, for it is the regular Tonic contraction of FOFOYET and ZOFONET, as cube is the Attic of FOFELOGN. The Lacedæmonians employed the ſubſtantive rug or TO x to fgnify any violent effort or impulſe, according to Plato (4) ; but it is probable that they themſelves ſpelt the verb, in their own dialect, with the h iuſtead of the F or T, LOB; whence we have, in Heſyehius,

600 VE %. ia A dns ebene Hero bs J, K 226:

(3) Lace, fro. Sophoc. Ajad. maſtig. Vi, 1414. 44) Cratyl. p. 412. Ed. Serr.

ZOBEIN, du,, rp K. OT, it, re, &c. together with LOYEO, TOT TAI, EOYEQEN, £OQMHN, and ATIOLOBEI ; all explained to the ſame purport, ſo as to appear evidently different forms of the ſame "ts written according to the different modes of different dialects.

The cauſe of the firſt adjectiveꝰs being written with the O inſtead of the 

A was probably the coaleſcence of theſe two vowels into the Q, in the de-

_ rivatives ZQTHP and ZQTEIPA : but this coaleſcence is probably of no

very remote antiquity, the old words being, according to the regular courſe of analogy, ZAQTHE/ and ZAQTIE, as appears from the Veletrian In-

ſcription (1). The verb ſeems once to have been written ZAOFQ, "hone

the Aoriſt «oawrw, or EZAOFZA.


HEN we conſider che fluctusting and uncertain ſtate of the Greek

pronunciation and orthography, prior to the Macedonian Conqueſt,

which made the Attic dialect the general criterion of purity and correct- neſs, we ſhall not wonder that this kind of confuſion ſhould have crept

into the compoſitions of an author, almoſt cozval, if not (as ſome ſup- pooſe) anterior to the general uſe of letters among his countrymen, For, though the poet and the orator are the poliſhers, the methodizers, and al- moſt the modellers, of language, it is to the grammarian and verbal cri» tick that their fine-wrought forms and dazzling colours owe their perma- neney, as thoſe of the painter often do to the chemiſt and varniſher. Prac- tical eloquence was a ſcience- regularly taught among the Greeks even be-

(1) EQTEIPA occurs as a title of Diana on the braſs coins of Agathocles; but upon

more antient ones, of Tarentum, &c. the initials of the ſame title are TA.

Pp. ; * ore


- och 4 I


fore the Trojan war (1), as being the only means of government where the rights of the governors and governed were wholly unaſcertained ; and every chief poſſeſſed juſt as much power as he could perſuade the people to allow him ; but the theory of ſpeech, or ſyſtematic grammar, was never regularly treated as a ſcience till under the Macedonian kings; when, one dialect being recogniſed as the ſtandard, men had a given point, from which they could meaſure the extent of every deviation, and trace the ra- mifications of every diſtant and obſcure connexion. Unfortunately, how- ever, this dialect was not the parent one; but, on the contrary, that which was moſt corrupted, or (as its admirers will ſay) moſt poliſhed, by local and cuſtomary peculiarities. Hence the antient grammarians, who conſi- dered this diale& as the criterion of purity, never explored the ſqurces of their own language, but endeavoured to correct the compoſitions of their moſt antient bard by the practice of thoſe who had imitated the very cor- ruptions which obſcured him. Great numbers of antient inſcriptions muſt then have exiſted, which, had they been examined, would have exhibited at leaſt the roots of his words in their genuine forms ; and from theſe their complete ſtructure might have been regularly traced, Few monuments of this kind have come down to us, and thoſe few have been too much neg- lected by Criticks and Grammarians. Nevertheleſs, the well-direQed la- bours of Hemſterhuiſe, Valkenaer, Damm, and Lennep, and, after them, of Villoiſon and Lord Monboddo, have diſpelled the clouds of grammatical jargon that obſcured the moſt important part of the Greek tongue ; that is, the flexions of the verbs. Thoſe who with to know the progreſs and detail of theſe great diſcove- ries will conſult the printed works of theſe learned perſons, particularly the Analagia Grace of Lennep. I ſhall here only give the reſult of them, in a ſhort table, ſhowing how the middle voice and the ſecond futures and Aoriſts have been formed out of different themes of the ſame verbs, only fragments of which have continued in uſe. Theſe fragments I ſhall place under their proper heads, and with the proper explanations, leaving

the ſpaces of all the obfolete forms, except the firſt, which is the theme itſelf, void.

% ee U. La 6 ; Though

N 3 2 g _ 4 lb... at,

  • Y wtf * ” . v \ ,

23 , * 1 A "© y 5 > a o * : K - = R G - „* -

— — Tre, the tic future ntracted TYTIEEN, ; Tron — — — TETY® ETETYSEIN

tracted contraſted fro 

— — —



ulgarly the 2 adopted for theſcontracted f hich has been Aoriſt. middle voice. TETYIAMAI, contracted in th ( The ſecond per-which ſeems o- ſame manner,

ſon ſingular is riginally to have 

contracted frombeen TETYTIE-

Aro to -AO Q1JKAMAI,  
A.  . 

5 a N N A

[ TY®OKEOMAITI : — — — TYTIHMI, the ETYTIHN, — N —— — active form uſedſvulgarly the ſe- in a paſſive ſenſe.ſeond Aoriſt. Tro, ETY®GHN, — — — the ſame, Jvulgarly "—_ 77 « Aorift, ( (

IN, TYMEEIN ulgarly the ſe- future cond Aoriſt.

. ? Be TYMATEELOA1 — TYTIEZOAI, TYTIZEZOAT, algarly the te. vulgarly the firſt »nd Aoriſt future middle,

TYTIEIZOAT, TYTIHZEZOAT, r TYTIEEE®A]), vulgarly the ſe- ulgarly the ſe- [cond future, ond future nddle. . — TYS@HEEEOAI Irma, — gariy t p ond Aor iſt, TYS9HNAT,

ulgarly the fir ont

[To face p. 106. 79 ö Infinitive, . * Preſent Perf, ſruture Perf. — 4 — — —— —

DYMEAL + — — 

- 5 — — — —


— * — k 3 — TETTOOAI, — TYTIZAZOAL, LOAF vulgirly the firſt Aoriſt middle, 1 ö ö




— TETYTIENAT, — . dopted for the

iddle voice. 

— — —




Though I have no doubt but that this hypotheſis is true, as far as it ſuppoſes the ſecond futures and Aoriſts and the middle voice to be moditi- cations of other tenſes and other voices, yet I cannot ſee any neceſſity for ſuppoſing the exiſtence of ſo many obſalete themes of the ſame verb as are here given, fince all the forms now extant may be deduced, by the regular licence of contraction, from two, TTT and Tren (with their re- ſpective terminations in · Mi), which are only variations of dialect, conſiſt- ing in the inſertion or omiſſion of the aſpitate. The ſuppoſing a termiua- tion in -E, in order to produce a future in -EEN, is not only unneceſ- ſary, but inconſiſtent with analogy ; for, as the termination in -Q formed the preſent perfect originally in -EKA, it muſt, by the ſame,rule, have formed the future in -E; which ſome contrafting by au elifion of the E, and others by an eliſion of the Z, it became -£Q and -EQ, corrupted to Q and -OY ; ſo that TYIEN, TTnEo, Tru, and TYIIOT MAI, are all the ſame tenſe, and from the ſame theme, which is the common one, TTIHTQ, the T being elided in the flexions. The regular future from a termination in -EQ muſt be in -HEQ or -EEEN, as the ſecond perſons ſingular from -OMAI and -EOMAI were originally -ELAI and -HEAT, changed by the Ionians to -EFAI and -HFAI, and thence conttacted to -EAI and -HAI, and ſtill further, by the Attics, to HI or .

Lennep ſuppoſes that the primitive form of the infinitive was the ſhort- eſt, that is, the Doric TT TEN, and that the other common and poetical forms (as they are called) are licentious variations and extenſions of it (1); but Lord Monboddo is probably right in taking the longeſt form for the original, TYITEMENAT, contracted by degrees to TTHTENAI, TTnTE- MEN, TTTEEN, TTnTEIN, and TTnTEN (2). In almoſt every word of the Greek we meet with contractions and abbreviations, but, I believe, the flexions of no language allow of extenſion or amplification, In our own, we may write SLEEPED or 8LEPT, as the metre of a line or rythm of a period may require; but by no licence may we write $LEEPEED.

Though the middle voice conſiſts of certain forms of tenſes belonging to the other voices, theſe forms were, at a very early period, employed to expreſs a particular meaning. To ſignify the doing of any thing in general

(1) Analog, Græc. p. 157. (2) Orig. of Languages, Part. II. Lib III. C. XIV. Fg without


without any particular reference, the active voice was employed ; but when it was done for the uſe of, or with a pointed reference to, the deer, the middle, as in

To pa Tor” 8x N, ] DN

Tlgwrov, entre ts va" dͤarec xoryr: poyrt

Nabaro 3” aureg Xup apurouro 3' aibona o. II. U. 228. and EZayaye wpopowg de, u mers ev awry;

Hyayro woes dur 1741 Tops prupics dove. Ibid. 188 & 190.

In an Athenian law, cited by Æſchines (1), the expreſſions of which muſt of courſe be preciſe and accurate, we find that MIr²⁵ nA ſignified the perſon who hired out, MILON.EAMENOE, the perſon who hired in, or for ' himſelf, and MEMIxenaukNOE, 7Zhe perſon who was hired, This I be- lieve to be the proper uſe of this voice ; for, when it is employed recipro- cally or neutrally, the n ſeems to be underſtood; wherefore, it is in fact actively.

Though the Greek tenſes are thus ſimplified, and reduced to the general principles of rational grammar, which prevail alike in all languages, it is no eaſy matter to aſcertain their preciſe meaning, and ſtill leſs fo, to ex- preſs it by the complicated auxiliary verbs, which the ſtubborn inflexibility of modern dialects has obliged us to adopt.

Dr. Clarke's note upon it is ſpecious and ingenious (2); but he has eluded rather than ſolved the difficulties, by giving his examples only from neutral aud paſſive verbs, and thoſe too in Latin. His ſtatement is Cas a os ſeen he was going.

coenabat, he was at ſupper.

fe action Fa upper. Time is wa 4 ædificabatur, it was in building.

of a perfect abierat, he was gone. action. Hage be bad ſupped. ædificatum erat, it was built.

(1) Kara TN (2) II. A. 37.



abit, be is going. it 2800 15 he i 14 at uppen. 01 Tot Li ædificatur, it. is building. 4

ſof an imper-

fect action 

Time preſent /

coenavit, be bas ſupped. 

1 ædiſicatum eſt, it is built, ſor an! mper- abibit, be will be going. fe action ccnabit, he will be at fupper.

Lædificabitur, it will be in building. 

— he will be gone.


or & perkect 0 abiit, he is gone.

Time future of a perfect

action cœnaverit, he will have ſupped.

\ ædificatum erit, it will be built,

Lord Monboddo has amply expoſed the defects of this ſcheme, oy given one of his own more complete; but, I fear, not much more Tatiefac- tory, It is as follows (1): moment; on Active.

FPA®N, Þ rwrie.

EPPA®ON, 1 was writing,

pana, I. ball or will write.

ETPAUZA, I wrote, or did write.

TErpAOA, I have written.

TEFPA®NE ELOMAI, TI all have written. 

EFEDPA®GEIN, I had written,


IPA®OMAT, I am in the act of "RM written. ETPA®OQMHN, I was in the act of being written. TPAHTOMAI, I all be 1oritten. TIT EFPA®OHN, I was written.

TETPAMMAl, I have been written. TETPAUZOMAIL, I ſhall have been written. ETETPAMMEHN, I had been written.

am to this hypotheſis, the perfect participle paſſive ought to ſig- nify that which has been done; but, nevertheleſs, TETEAEEMENON

(3) See Orig. of Lang, Part II. Bock I. Ch. XII. EETAI


ELTAI does not mean ſball have bren\finiſhed, but ſhall be finiſhed completely, It is difficult to conceive how an action can be complete, aud yet preſent ;

ſince the very completion of it\renders'it pas: but, nevertheleſs, this ſeems to have been the ſenſe of the tenſe which is commonly called the præter- perſeft, but which Clarke more properly calle the preſent perfect. The pluſquam perfetlum, or paſt pere, ſeems, in like manner, to have been often uſed to ſignify the fuddeneſi of the action without having reference to

an event completely pas, as in reg, 09109 6006 Idyw—nerqnxe 3 49. Pecyxu and eErEmuy in many inſtances, Lord Monboddo, indeed, ſuppoſes theſe to be the preſent imperfecis of new themes dd, rrrpyxu, GCN, &c.

formed from the preſent perſeds tenſes (1) ; and it is certain that ſuch new themes were occaſtonally uſed, but, 1 believe, not ſo frequently as that

learned writer imagines, for Homer's narrative, when delivered in his own perſon, is always in a paſt tenſe (2); and it is rather ſingular that, if he choſe to deviate from his general practice in this reſpect, he ſhould have done it only when employing theſe augmented forms, and thus introduced a licentious enallage of tenſes, which he never allows himſelf on other oc- caſions, merely to introduce a ſet of licentious or irregular words. We may obſerve too, that did is certainly a paſt perfect in Il. M. 37, and like- wiſe in Il. K. 187, otherwiſe it could not accord with the ſucceeding verb riręaparo, as the ſenſe requires it to do. The learned Judge has, indeed, turned eana\&ro and ge, Which are equally: paſt perfects, into Aoriſts ; and, by the ſame licence of tranſmutation, he might have made one of Trrpapero, or, indeed, of any other form; but he ſhould have recollected that Herodotus, an author whom he profeſſes to have ſtudied ſo accurately, employs undoubted paſt perfeAs in exactly the ſame ſenſe as Homer has theſe diſputed forms— avrog aviCeCmery nav nat” αννον Hege twy α. —then he himſelf ſuddenly went up, and others of the Perſians proceeded after him (3). Dr. Clarke has accordingly underſtood all theſe forms to be paſt perfedts, though their meaning does not — OE with the uſe of

. (1) Origin of Languages, Vol. II. p. 187. (a) I would here be underſtood to diſtinguiſh between narvative and 4 eſeription ; for de- feription may properly be in the preſent tenſe, ben the narrative to which it belongs is in the paſt, . (3) Lib. I. S. 84.


ON THE GREEK GLPHABET. a1 that tenſe in modern l and his opinion certainly does not merit the contempt and aſperity with which it has been treated by the learned Judge; for it is juſtified in this iuſtance by the very high authorities of Virgil and Horace, both of hom under ſtood Homer's expreſſions exactly as Clarke has, and thought them beauties worthy of being tranſplanted into their owt! language. The former has, fic fata, gradus evaſerat al. 4 1), CE correſponding with &s 1K latter, . "As hoe prementes verterant bis mille d; ns

1 10 Gal canehtes Cafarom, parallel with. u. wt Greim mr iti

— — — 7 d 14 — NCe Io anbei .

1 „ . » „ - * 4 v = FF 7 1 V " j - . 1 w - ” ' 92 2 p " 1 > 0 "© 1 8 . . 4 . i - * 5

  • E learned reader muſt have obferved that, in the whole courſe of

this Enquiry, I have tacitly rejected the evidence of ſome very cele-

brated and important monuments of antiquity, firſt publiſhed in the Me- moirs of the French Academy of Inſcriptions and Belles Lettres, and fince cited as authentic by every writer upon thĩs ſubject. I mean the inſcriptions ſaid to have been diſcovered in the neighbourhood of Lacedæmon, by the

Abbe Fourmount, during a journey through Cores, undertaken by order. of the late King of France.

M. Fourmont is ſaid to have been a porings hs Antiquary, withour

taſte or invention, but of immenſe induſtry and rigid exactitude in com- piling, and ſo devoted to antient learning, that he underſtood Greek and Hebrew better than his native Frencl (2% Of his proficiency in the two

(1) En. IV. 68 8. (gz) Recherches fur les Arts, Vat 1. latter

= - * ©, 2 *

— 1 * *

1 ** 7

3  ' 

\ %


  • 1


latter languages, I am not a competent judge 3 but of bis (kill in the firſt,

1 may perhaps be able to give the reader a juſt idea, by a free and candid

examination of the inſcriptions which he produced. This examination 1 feel it iucumbent upon me to make, as an apology for my preſumption in differing in opinion with ſo many of the ſirſt ſcholars of the age, who have quoted theſe inſcriptions as Wanne (peciqppus of the d antient writ» ing extant.

When Mr. Fourmont returned from Greece, be LES out that & had made vaſt diſcoveries, having got an anticyt copy of the laws of Solon, and, by employing two thouſand men to dig in the ruins of Amyclæ, found written monuments of much more remote antiquity than any that had hitherto been produced, Specimens of theſe he publiſhed in the year 1742; but from ſome cauſe or other did not proceed, but left his manu- ſcripts in the King's Library, from which other ſpecimens have been ſince publiſhed by the Benędictines in their Trait“ diplomatique, and by the Abbe Barthelemi in ſucceeding volumes of the Memoires of the Academy. Theſe, however, form but a ſmall part of the collection, the reſt being, as Count Caylus ſays, withheld from the publick on account of the expence neceſ-

ſary to make engtayings of ſuctCa number and variety of characters as are 

contained in them. A large volume of manuſcripts, copied from Four- mont's originals, under the direction of the Abbe Barthelemi; is, indeed, now ſhown in the Library but it cannot be to theſe that the Count alludes, for they contain very little variety of character, being chiefly mutilated and incorrect copies of inſcriptions already publiſhed, The originals, however,

of theſe are not ſhewn, any more than of the very curious and important

ones publiſhed ; and as for the laws of Solon, they are now given up, as well as the two thouſand men employed at Amyclæ; it having been diſco- vered that the whole Peloponneſus would ſcarcely have afforded ſo. many. Fourmont, indeed, did employ all that he could collect, not in diſcovering inſcriptions, but in breaking to pieces thoſe previouſly diſcovered, that fu- ture travellers might not detect his errors and frauds (1).

When ſo intelligent and experienced a perſon as Count 8 talks of

the enpencſ as the great impediment to pubparen, we cannot but ſuſpect

(x) Of this I was informed by the late tr. Stuart, who followed 6

1 RN Os 8 that 


that he adduces ſo frivolous a: reaſon, merely to cover a more ſolid one,

which he thought proper to ſuppreſs out of reſpe&t to the Academy; for

he muſt have known that the expence of engraving or caſting all the dif- ferent variations of character of which the Greek Alphabet is ſuſceptible,

could ſcarcely be. an objc & of importance to an affluent individual, and

much leſs to an illuſtrious public body, or powerful prince. Perhaps the

fair and free examination of thoſe already publiſhed, which 1 ſhall here

give, may bring to light the concealed. reaſon for withholding the reſt. The authority of the Academy, uuder which they were firſt uſhered in- to the world, has hitherto prevented any ſuch examination from taking place; otherwiſe, I am perſuaded that ſuch men as the authors of the Traite diplomatique, Abbe Winkelmann, Mazochi, M. Auſſe de Villoiſon,

and the preſent Biſhop of Cheſter, would never have quoted them as au-

thentic ; for as to. the character of Fourmont, and his want of invention and ingenuity to compoſe ſuch forgeries, they are but poor palliatives at

beſt, and will, I think, loſe the little efficacy, which they might otherwiſe f

have, when we become acquainted with the exact degree of theſe qualities requiſite for ſuch compoſitions. The author of the Recherches ſur les Arts has, indeed, adduced ſeveral other arguments in favour of them, the prin- cipal of which will be hereafter conſidered. His reaſon for undertaking a formal defence of them, was to anſwer objections which 1 firſt put toge- ther for his uſe, and which I now re-ſtate, nearly in the ſame form, and ſubmit to the judgement of the Learned ; only intreating every perſon who ſhall again differ with me in opinion, and think my remarks worthy of animadverſion, to make the reply generally to them all, and not, like the learned author abovementioned, oppoſe a profuſion of argument to thoſe parts which appear weak and harmleſs, while the reſt are left, wehen and unnoticed, to prey upon the ſpoils of the Academy.

he inſcriptions publiſhed contain ſpecimens of writing from the earlieſt period of fabulous tradition down to the ſubverſion of the Greek Republicks from Eurotas, a king ſuppoſed to have reigned in Laconia ſeven. genera- tions before the Trojan war (1), down to Philip of Macedon. In mouu- ments, engraved at periods ſo remote from each other, we might expect to

(1) Pauſan. Lib, III.



HL. ( . FT * . Aion

. 4


find great variations both in the form and uſe of the letters; but, never- theleſs, they are ſo nearly the ſame as to appear of one hand-writing, and of one perſon's compoſition, We have the terminations of names in the oblique caſes the ſame as in Pauſanias; and all the barbarous forms of let- ters, ſuch as the (Sigma and E Ep/filon, employed under the later Roman Emperors, The Sigma in the earlieſt inſcriptions is, indeed, taken from the very antient medals of Gortyna, in Crete, upon which we find the word SNVT%O1, which Fourmont, like ſome other Antiquaries of equal ſagacity, took for FOPTYNE ; whereas it is FOPTYNI, the abbreviation of FOPTYNIQN, found upon other medals of the ſame city; the Iota be- ing of this form, as before obſerved, on the medals of weeks Poſidonia, and in the Veletrian Inſctiption.

This remarkable conformity has been attributed to the pertinacious ad- herence of the Lacedæmonians to their antient manners and cuſtoms ; but it ſeems to have been forgotten, that theſe manners and cuſtoms were twice totally changed during the period comprehended in theſe inſcrip- tions; firſt, by the invaſion of the Dorians, and, afterwards, by the Inſti- tutions of Lycurgus; and that, in the age of Homer, or, at leaſt, in that of which he writes, which was conſiderably later than the earlieſt of theſe inſcriptions, Lacedæmon was the ſeat of n and luxury inſtead of arms and diſcipline (1). The forms of the bucklers alſo, upon which two of the inſcriptions are engraved, are totally unlike the fimple round ſhields of the antient Greeks, or indeed of any other antient people, they being in abſurd fanciful ſhapes, wholly unadapted to the purpoſes of defence (2). The mode of writing the titles of the magiſtrates too, in larger letters than thoſe employed in

their names, is without example in any genuine monument of antiquity that I have ſeen (3); and it is obſervable, that one of the ſtones is repre- ſented as broken in fo artiſt-like and regular a manner, that it could not have been the reſult of accident (4); for, if ſo many fractures had been cauſed by the fall of ruins or the decay of time, the edges would neceſſa- rily have been ſplintered or corroded ſo as to deſtroy many of the letters. I ſhall, however, waive the conſideration of theſe ſuſpicious peculiaritics,

(1) Odyſſ. 4, (a) See Pl. V. VIII. & IX, See PL VI. & VII. (4) See Pl. IV.


as well as the ſingular forms of the ſhields and letters, becauſe whim and caprice might have operated in antient as well as modern times: but errors in orthography, grammar, and dialect, the blunders of dictionary- makers, tranſcribers, and editors, transferred into monuments attributed to remote antiquity, will, I flatter myſelf, if , be deemed of themſelves ſuf- ficient evidence of impoſture,

The moſt antient of theſe monuments. is a tewple or chapel dedicated to the goddeſs Onca or OG a, which Fourmont pretended to have diſco- vered, but which no other traveller has been able to find, notwithſtanding the maſſive and almoſt immoveable ſtability with which he ſays it was

built. As this chapel is ſuppoſed to have been dedicated in the time of the King Eurotas abovementioned, the father-in-law of Lacedæmon, from

whom the city derived its name, it was neceſſary to find. ſome other title for the Lacedæmonians in the dedicatory inſcription, Meurſius eaſily ſup-

plied this deficiency ; for in the text of Heſychius, as it then ſtood, he

found Ixrerxpare; Aaxwe, whence he concluded that Iertusgareig, Of Ine TWxpaTe, was an antient name of the people of Laconia(1). Fourmont, therefore, adopted this name with a whiaſical alteration, and gave as the yotive inſcription of his chapel, OPAL IKTEPKEPATEEZ (2). It has ſince, however, appeared, that this name is merely the creation of a blun-

dering tranſcriber, who transformed two verbs, the one explanatory of

the other, into a ſingle noun ; ſo that, inſtead of Ierexpurits Aaxwveg, WE ſhould read weriv” xpare:; Aaxuves, as the laſt editors have juſtly obſerved.

Thus, by a ſucceſſion of error and impoſture, a fabulous perſonage of an-

tient tradition has been made to anticipate the blunders of a tranſcriber, committed in copying a dictionary-maker of the third century of Chriſtia- nity (3) ; by which means the French academicians have been. enabled, not only to call into being a people that never exiſted, but alſo to fix the date of their dominion in the Peloponneſus as readily and accurately as that of the Franks and Normans in their own country (4).

(1) Miſcell. Lacon. Lib. III. C. VIII, (aa) See Pl. III. Fig. 1. (3) See Fabric. Biblioth. Græc. lib. IV. c. XXXV. Though the original author ſeems:

to have flouriſhed about that time, his work has been mutilated. and interpolated by later hands,

W See Mem, de E Acad. t. XXIII. p. 415: Q 2. The


The next inſcription is a catalogue or chronicle of the prieſteſſes of Amyclz, beginning about the ſame time; Laodamia, the grand-daughter of Eurotas, being the third prieſteſs in the liſt (1). By a peculiarity of idiom, theſe prieſteſſes are called MATEPEE KAI KOYPAI TOT ATIOA- AONOE ; titles, for which neither M. Barthelemi, nor the author of the Recherches, &c. have been able to produce any authority (2), though they ſeem both to have wandered over the pages of every book extant in the -Greek language. The latter has, however, incautiouſly ſuggeſted the cor- reſpondeut titles in the modern French convents of nuns, which afford a clear and undoubted illuſtration, Les MerEs ET LES FILLES DU BON Dixu were+familiar to Fourmont's mind; and he not only adopted the idea for his antient IKTERKERATEANS, but, by a refinement of inconſiſt- ency and abſurdity, made them expreſs it in all the crudity of its native idiom.

In reading the names we find other peculiarities of idiom not leſs extra- ordinary, ſuch as APILETANAEPO, APIEETOMAKO, KAAIKEPATO, TEkKEnAO, EEKOAA, LEKIAO, and EEKENOMA ; which, I ſuppoſe, are intended as a ſort of Hebræiſms, modelled upon the ſame plan as IK-

TEPKEPATEEZ for IKTEPKPATEEL ; theſe names being the genitive 

caſes, according to M. Fourmont's declenſion, of words, which, in ordinary Greek, we ſhould write APILTANAPOE, APIETOVIAXOE, KAAAIKPA- THE, EKEIAY, xkOAAZ, EKIAAOE, and EXNNOMAE. This learned gentleman had, it ſeems, received as incontrovertible truths, the wild opinions, or, as Lennep calls them, the /plendida deliramenta, of thoſe Criticks, who, ſoon after the revival of literature, endeavoured to deduce the Greek from the Hebrew, and other oriental roots. He alſo knew from Joſephus (3), that the Lacedæmonians and Jews looked upon themſelves as ſprung from a common ſtock, aud cloſely allied by the ties of conſan- guinity ; whence he naturally concluded that Hebræiſms would be more likely to occur in the writings of that people than in thoſe of any other

(1) See Pl. III. Fig. 2, I. 3.

(a) See le jeune Anacharſis, vol. I. p 509, 4to ed.; Recherches ſur les Arts de la Grce, vol. II. p. 251.

(3) Antiq. Jud. Lib XII. c V. & Lib, XIII. C. IX.; and Meurſ. Miſcell. Lacon, Lib. . C. VII.


ON THE GREEK ALPHABET. 117 Greeks; and it muſt be owned that; in theſe inſcriptions, he has given them a large ſhare, ſo as utterly to ſubvert the analogy of their own lan- guage. Unfortunately, however, the Hebræiſms which be has attributed to theſe fabulous chiefs of the Peloponneſus, who lived (if they lived at all) fifteen centuries before the Chriſtian æra, are the Hebraiſms of the Maflorethic criticks, who regulated the pronunciation of that language, by adding the vowel points to the text of the Bible, ten centuries after it. The flexions of theſe words are not leſs whimſical and extravagant than their conſtruction. FEKETAO is terminated according to the old Holic, or (what is the ſame) the very antient Doric; but TEKEAA and ZEKE- NOMA are, according to the later Doric, which was either -poſterior to Homer, or not known in the country where he compoſed, as no inſtance of it occurs in either of his poems. Theſe names, in the inſcription, ate immediately ſubſequent one to the other (1), ſo that the variation could not be intended to mark any revolution, as ſome other changes of ortho- graphy are, which will be duly conſidered. KAAIKEPATO would have remained 1nexplicable to me had not the author of the Recherches expoſed the blunder by participating it. The genitives HEPMOKPATOE and NPOKONEZTO, in the Sigean Inſeription, he ſays, are alike abbreviations or corruptions of the regular genitive termination in -O (2); by which it ſeems that this ingenious author (who has certainly ſhewn great acuteneſs and ſagacity in explaining monuments of art) took his notions of the de- clenſions from Fourmont's inſcriptions, who has confounded the two clafſes of nouns terminating in -H, which are vſoally, but improperly, called ſimple and contracted, for both are alike contrafted, though the primitive extended forms of the oblique caſes are leſs frequent in the firſt than the ſecond. As imitators generally copy their originals in an inverſe ratio of their merit, that js, by adding as much to'their faults as they loſe of their merits, the author of the Recherches has added another claſs of

nouns, namely, the adjectives in OS; to ſwell the con fuſion. Fourmont

Having feen that names compoſed of KI THE formed their common gem- tives in Or, and that this diphthong was repreſented in very antient in-

  • the ſingle O, concluded that words n of KPATHx

(1) See Pl. IV. I. 13 & 14. ä (2) Vol. * p. 213 & 223.

bs . were 


were liable to the ſame variation, and therefore wrote KAAIKEPATO in- ſtead of KAAIKPATEOE, KAAIKPATOTE, or KAAIKATOx, either of which would have been Greek. , This error is ſo groſs, that, were it not perſevered in through the remaining inſcriptions, and illuſtrated by Meſſrs. Barthelemi and D'Hancarville, 1 ſhould have ſuppoſed it to be an error of the engraver : but we have EYPIKPATEO, AAKAMENEO, and KAEO- MENEN, all upon the ſame principle, for EYPYKPATEOZ, AAKAME= NEOE, and KAEOMENEAE,

Nothing expoſes ignorance ſo effectually as an anſvcpelafpl attempt at ſcientific accuracy. To mark the period of the Dorian invaſion under the Heraclidæ, the terminations. of the names af the prieſteſſes are changed, from what Fourmont thought folic or Ionic, to Doric. Hence AMY- MONEE, in the beginning of the inſcription, becomes AMTYMONA after- wards; but the ending of feminine names in two Epſilons inſtead of an Eta is unauthoriſed by any autient monument, and expreſſly contradiated by a paſſage of Plato (1). Neither would this orthography, if juſtified, exhibit the diale& of the antient Laconians, which muſt have been the old folic; for Strabo expreſſly tells us, that the fugitives, who quitted the Peloponneſus under the deſceudants of Agamemnon, when invaded by the Dorians, were the founders of the firſt Aolian colonies in Aſia (2).

M. Barthelemi, the editor of this Amyclæan Chronicle, thinks that the beginning of it, comprehending the names of the firſt ten prieſteſſes, has been renewed like the ſecond Sigean inſcription, but that all the other names, together with the dates of their adminiſtration, were inſerted in order as they ſucceeded to each other; ſo that this chronicle, when entire, muſt have given the dates of all the great events of the fabulous Hiſtory of Greece; for as Laodamia, the daughter of Amyclas, is the third prieſteſs on the liſt, the time of his reign. muſt have been correctly aſcertained, and, of courſe, that of his immediate progenitors, Lacedæmon and Eurotas. This wauld have led to a knowledge of the time of Cadmus's arrival into Greece, of the ſieges of Thebes and Ilios, of the return of the Heraclidæ, and all the other diſtinguiſhed, events of poetical tradition, the dates of which none of the great writers of antiquity could fix with any degree of

() Ov yap H xda, a E ro waa. Cratyl, (2) Lib. XIII. p. 872.


ON THE GREEK ALPHABET. 119 probability. This monument, therefore, though exiſting (if it exiſted at all) in one of the moſt celebrated temples, and moſt frequented provinces, of Greece, muſt have eſcaped the notice of all the inquiſitive travellers and diligent Antiquaries, who, during ſeveral ſucceſſive ages, endeavoured to rectify antient chronology, 3

I know that arguments ſimilar to theſe have been urged againſt the au- thenticity of the Parian or Arundelian Chronicle; a monument, which the ſurface alone proves to be undoubtedly antient; for no chemical pro- ceſs can produce the ſtains, corroſions, and calcareous concretions, Which mark that marble. It muſt be remembered, however, that this Chronicle contains only the private opinion of one of theſe conjectural chronologers,/ and probably of oue not in the higheſt repute; wherefore, we need not wonder that it is not cited by any antient author. But the Amyclæan Chronicle, if genuine, muſt have afforded undoubted evidence, as far as it went; for though the events which it directly aſcertained might have been but few, yet theſe few would have ſerved as points of obſervation, from which the bearings and diſtances of many others might have been diſco» vered. In its preſent ſtate, the Abbe Barthelemi has aſſerted that it can be of no ſervice to chronology ; but M. D'Hancarville has thought differ- ently, and, in a long Commentary upon it, proved that it fixes the reigns of the fabulous kings of Lacedæmon to the period in which Lydiat and Marſham, after the Parian Chronicle, had placed them (1); as, indeed, it naturally would do, it having been fabricated from their where 42 and thoſe of Cragius and Meurſius.

According to this calculation, Eurotas and Lacedæmon were contenþo- raries with Cadmus, to whom general tradition has attributed the intro- duction of letters into Greeee (2). If this tradition be well-founded, Eu- rotas could have written in no other character than the Phœnician, ſuch as we have ſtill upon the very antient coins of that people, and their co» lonies in Africa, Spain, and Sicily, Theſe characters, as is well known, were ſixteen in number, written from right to left, and moſt of them very different in form from thoſe of the Greeks (3). Herodotus, however, ob- ſerves, that the Cadmean letters upon a tripod dedicated by Laius, the

(1) Recherches ſur les Arts, &c. vol. II. 5 (2) Ibid. Lib. II. C. II. p. 333.

(3) See Dutens, Diſſert. ſur les Medailles Phœnic. & . * at her


father of OEdipus, which be ſaw at Thebes, differed but little from the Ionian (1): but whether the letters were changed between the age of Cadmus and that of Laius, or whether the juſcription ſhewn to the Hiſto- rian was a forgery, is uncertain. The Ionian letters on the medals and other monuments of his age, now extant, are evidently very different from the Phoenician ; and as for thoſe upon the temple of Oga, they differ eſ- ſentially from both, being written from left to right, and having the Omicron triangular like the Delta, and the Rho like the Alpha (2), only turned the other way; which are forms alike unknown to the Phoenician and lonian alphabets. The other letters, both in this Inſcription and the Chronicle, are like the Ionian fantaſtically diſtorted.

I am willing, however, to abandon this ſtrong argument againſt the authenticity of theſe monuments, and to admit that letters were known in Greece before the introduction of the Phoenician alphabet by Cadmus; for which my principal reaſon is, that the firſt piratical ſettlers, who brought letters from Greece into Italy, brought an alphabet much leſs perfect, and therefore, probably, more antient, than the Cadmèan. That of the Eugubian Tablet contains only twelve ſingle letters, unleſs the Vau is to be reckoned diſtin from the U, with which Gori joins it, as being the aſpirated U (3). Theſe are, probably, the original Pelaſgian letters, as firſt brought into Italy ; for, without admitting the conjecture of Gori, that this inſcription was engraved two generations before the Trojan war, we may ſafely allow it to be more antient than any other written monu- ment extant.

The Pelaſgions are ſaid to have been the firſt coloniſts who ſettled in Italy after the Tyrrhenians (4); and, according to Pliny, brought letters into Latium (;). In this, however, he ſeems to have been miſtaken, for the Latin letters, as well as language, are clearly derived from the ZEolian or Arcadian (6), which were nearly the ſame as the Cadmean, and had ſeveral characters of which the Pelafgian alphabet of the Eugubian Tablet is deſtitute, There is, however, a reſemblance between the forms of the reſt, from which we may infer that they were originally the ſame, and

(1) Lib. V. C. LVII. (2) See Pl. III. Fig. 1. (3) Proleg. ad Tab. Etruſc. (4) Dionyſ. wpmy» VI, 347- (5) Hiſt. Nat. Lib. VII. C. LVL (6) Ay Lib. I. C. VI, & Corinth, Tips dν,ů)P&ꝙ̃.




only varied as they advanced in improvement (1). The Latin are fard to have been introduced by Evander from the Peloponneſus about the time of - the Trojan war (2), and were, without doubt, ſuch as were in uſe in that vountty in that age. Their number was then fmall; but the Romans continued to add to them until they produced the alphabet now chiefly prevalent in Europe. The Pelaſgiau, probably, came into the parts of Italy weſt of the Tyber at a much earlier period. The Eugubian Tablet has no B, G. D, or O; the three firſt being included in the correſpondent mutes of the ſame organs, and the laſt in the U, which being employed as a confonant, or rather aſpirate, formed the Pelaſgian Jun, the Romau V. and our W (3). This letter is generally called the Phoenician Vau 3 but, I believe, it is not to be found upon any authentic monument af that people; whereas in the nes and Etruſcan inſcriptions it occurs per-


Whether theſe anticnt ieee tend hats letters ho the Phœni- EE 

cians at a period anterior to the expedition of Cadmus, or whether both the Phcenicians aud Pelaſgians received them from the Aflyrians . (whony Pliny mentions as the inventors of writing (4) ), or from ſome people ſtill more antient, is impolhble now even to conjecture. The Pelaſgians appear unqueſtionably to have been the firſt people of Europe among whom arts and letters were at all cultivated (5); for as to the traditions mentioned by Strabo, of the antient ſplendour and civilization of ſome nations in Spain, they are unſupported by the teſtimony of any exiſting monuments, and, therefore; probably fabulous (6). The Athenians derived their origin from the Pelaſgians (7), who are ſaid, by Ephorus, to have founded the Oracle of Dodona (8), the moit antient in Greece; and which; by the account given of it by Homer, ſeems to have reſembled thoſe of the Druids (9).

They were ſpread over all Greece, and part of Aſia; and it is probable that moſt of the tribes, mentioned by Strabo and Pauſanias, as formerly occu- pying different parts of the Peloponneſus, were only clans of this people; 6 n to n the whole rannten was antiently called * 1

« F See Gori ind c. 62 Dionyſ. Halic. Antig. Rom. lib. I. (3) See Foſter on Accent and Quantity, c. IV. (4) Hiſt, Nat. 1. VII. c. L VI. (5) See Strab. lib. VII. (6) Ibid. lib. III. (7) Herodot, J. VIII. c. XIV.

(8) Apud Strab. lib. VII. (9) 11, u. 234+ F

 5  5 laſgia; 


laſgia ; and we know that the ſame language prevailed and continued through every part of it until the Ionian and Dorian invaſions (1). They came into Italy from thence (2) ; but at what time cannot be aſcertained. It was, however, between the arrival of the Tyrrbenians and that of Evander. ; but when the Tyrrhenians arrived is quite uncertain, for Dio- nyſius of Halicarnaflus very prudently rejects the ſtory of Tyrrhenus, the grandſon of Hercules (3), whom we may ſafely rank among thoſe imagi- nary heroes, who were called into being to account for the name of a country; and, I believe, if we add Eurotas, Lacedæmon, and Amyelas, to the ſame liſt, we ſhall do perfectly right, notwithſtanding the pretended cozval inſcriptions which bear their names. At all events, the expedition of the Pelaſgians could not have been anterior to the period in which theſe princes are ſuppoſed to have reigned ; for the Grecian ſea was then, and for a long time after, poſſeſſed by the Phoenician and Carian pirates, who, having ſettlements on moſt of the adjoining iſlands, muſt have reſtrained the inhabitants of the Continent from making any conſiderable naval expe- ditions (4).

Minos, King of Crete, was the firſt of the Greeks who acquired a naval power, and opened the ſea for his countrymen, by expelling the Phoeni- cians and Carians from the iſlands. They then became pirates in their turn, and extended their predatory expeditions all along the coaſts of Afia and Italy (5). Minos, according to Homer, was two generations before the Trojan war, his grandſon Idomeneus having been a leader in it; but, as he was then advanced in age, we may, in calculating, allow Minos to have been three generations before the war, which will Mill place him four generations later than the fabulous king Eurotas.

If the Pelaſgians could not have come from the Pelopenneſits into Tuſ- cany before the ſuppoſed time of Eurotas, it naturally follows, that the al- phabet which they brought with them could not be more antient and im- perfect than that then in uſe there. But how does this accord with the votive inſcription attributed to him, where we find the F in the common Ionian form (which is that of the Pelaſgian II), and the O diſtorted into the form of the 446), whilſt neither of theſe letters exiſt in the Eugubian


(1) Strab, lib, VIII. (a) Dionyſ, wwe. 348. (3) Antiq. Rom. lib: 1. (4) Thueyd. tib. l. (5) wid. (6) See Pl. III. Tig · l. 1 5 : inſcription ?


Iaſeription The reſt are Ionian chatacters variouſly diſtorted, and written from left to right ; whereas both the an and Feen wrote from right to left. Theſe inſcriptions, e appear to be falſe, ach hypotheſis 19 we adopt, that of the Cadmèan being the primitive Were of n or that of the Pelaſgian having preceeded it. The next monument that offers itſelf to our e ee is one of the votive ſhields abovementioned, upon which is inſcribed the pedigree of Teleclus, King of Sparta, who is ſaid to have reigned early in the eighth century before the Chriſtian æra (1). This pedigree is taken exactly from Meurſius (2), except a trifling variation in the ſpelling, ſuch as a K for a X in the name APXEAAOE, Where the author found ſuch a genitive caſe; as AABOTAE is difficult to gueſs, unleſs he copied ſome error of the preſs, as I am inclined to ſuſpect. The word BATOx for ADOZE, à leader, he might have got from Cragius or Meurſius, who took it from an erroneous; or interpolated. paſſage of Heſychius, who firſt explains it to be xxacope eprys, adi, a fragment of a loaf or cake, which is right; for ayw or a vun, to break, was, as before obſerved, written with the F, which the Laconians changed to a B. He afterwards adds xa Baoinws xa, SRaTIT H, and a ling and a common ſoldier, which is certainly erroneous, and probably, interpolated ; for a, to lead, appears always to have been begun with a vowel, and many of the explanations in Heſychius are of later date than the original work, and of no authority. Fourmont, however, was not {killed in criticiſm, and therefore took every thing for granted which he found in the Dictionary, that common oracle of dunces. The next inſcriptions, according to the order of their pretended * are two tables, containing liſts of the kings, ſenators, and magiſtrates, of , Sparta, during the celebrated Meſſenian war, which employed the arms of that Republick during a conſiderable part of the eighth century before the Chriſtian #ra (3). Lo commemorate the events of this war, M. Fourmont and his commentators think theſe inſcriptions were engraved ; and indeed we know of noother purpoſes for which they could have been engraved. There is, however, no mention of the war, or any thing elſe in them, but merely '

(1) See Pl. V. 7 (a) Laconic, i in Grec, Thel. Antiq, (3) See Pl. VI. & VII. n R 2 the


the titles and names of magiſtrates, the former all taken from Cragius and Meurſius. As an excuſe for this peculiarity, he cites the known taciturnity and conciſeneſs of the Lacedzmonians ; not recollecting that this coneiſe- neſs conſiſted in expreſſing a great deal of meaning in a very few words, and not in employing many words to expreſs. no meaning, which is the caſe with theſe inſcriptions ; for, had there been no other memorials of the war, no one could have gueſſed that they had related to it; or, in- deed, that it had ever exiſted, Many of the magiſtrates could not have had any ſhare in it, as their offices were merely eivil; neither is it pro- bable that the faſtidious modeſty of the Lacedemonians would have ro- corded the names of thoſe who had; ſince they did not condeſeend to men tion a ſingle individual, not even Leonidas hinſelf, in the inſcriptions which recorded their noble ſacrifice at Thermopolæ.

In the titles of magiſtrates inſcribed, we find all the miſtakes of Cragius- and Meurſius exactly followed; ſome of which are confiderable, The for- mer, in ſpeaking of the "Apurga:, or rrgulators, had afferted; that they were called Af,‘ by Heſychius (1), whence: we find F MOETEPES: in the infeription : but the words of Heſychius are, *Agpogn;. 5 weh τéũʒ £T1jubAmTh; 815 UrmROAY TONW. Kat xe dv gg TW err TT; Pee iE Ape pegner; Auyorraus The latter part of this paſſage 18. evidently corrupt, and“ in its preſent ſtate conveys no meaning; but if, inſtead of aur, we' read vIw, as has been propoſed, the ſenſe will be, 1w6-/ones adapted 15 the ſbun- dation of the door-caſe are called 'Agubgznp;, The provincial governors of the Lacedemonians are always called *Apuoges z but it does not appear that: there were ever any regular domeſtie mapiſtrates of this kind; though. there might have been ſuch occaſionally elected to-controul. private man- ners. Theſe, however, were probably called *Apperove, a title of the ſatne- import; which, Heſychius ſays, belonged to certain magiſtrates of Sparta, choſen to regulate the conduct of the women. Fourmont, however, who ſearched no farther than Cragius, has put theſe down as another claſs of regular magiſtrates ; and, what is more extraordinary, put them down in a record intended to commemorate a War. In ſome inſtances he does not appear to have read more of his compilers than the heads of their chapters

) be Rep. Lacon, lib. II. c. XIII. 7 _ otherwiſe

ON THE GREEK ALPHABET. 125 otherwiſe he would not have given us ſueh a magiſtrate as the BOYATOP, who, according to all accounts, was only the head-boy of each claſs or eompany of the youths who were educated by the State. There were of courſe feveral of them at the ſame time; and as the waoopos, or public tutor (whoſe office Fourmont meant to ſignify, but miſtook the title), could not attend perfonally to all, he made theſe leading youths his depu- ties. That an antient Lacedæmonian ſhould have committed ſuch a blun- der as this, is as improbable as that an Etonian ſhould miſtake PROPOSTOR: for provosr.

Another extraordinary magiſtrate in theſe inſcriptions is the ANIOKA- PATHP; the y ivo x pρα²r , Or public ringe, of leſychius. The word is evidently carrupt, and ſhould be either qvoxparnys or q1oxapr1Cs probably the latter from iner *, and age, as has been conjectured. The Dorians would naturally have begun it with the A; but that A would, in the time of Theopompus aud Polydorus, have been preceeded by the aſpirate F or H in this title, as well as thoſe of the « de and Accu. The Digamma- would probably have been alſo employed, ſo that the regular word would Rave been FANIOXFAPTHE. The Lacedæmonians did, indeed, as before obſerved, employ the P for the 2; but they probably did 1 it regularly, and. notcapriciouſly, as it appears in the inſcriptions, and as Fourmont found: it in the oompilations of Cragius and Meurſius, gleaned from different au- thots, of different ages, and different dialects. Baro ought to have been written ATOP or BOYAFOP, BOYBAFOZ ;- ATEEIAAOE, BADEEIAAOZ 3; Kc. but Fourmont has written'the names as he found them in the books

which he conſulted, without conſidering this inconſiſtency, Even the name of the ſame perſon is written in the ordinary manner, ©EONOM-" Hor, in the one inſcription, and half Laconized to £IONOMIOE in the other. The name AkO is even Latinized ; for, 1 believe, in every dialect of the Greck it muſt have been AEQN or AEON, ©

The 1 for the T was employed in ſome inſtances by the Lacedæmonians, as Fourmont had probably heard; whence we have ſach words i in the in-- ſeriptions as NOMOOIAAKEZ and EYPIKPATEO, "which 1 ſhould have ſuſpected to be errors of the engraver or copyiſt, had not the termination of the latter made even a groſſer blunder perfectly conſiſtent with the learning and ſagacity of the author. This termination is evidently. a ſyſtematic, and



not an accidental, error, as it is perſevered in through- many words, and formally illuſtrated and defended by Meſſrs. Barthelemi and D'Hancarville; the firſt of whom very gravely tells us that it is Doric; and the latter, to corroborate his aſſertion, cites the word *'EPMEQ from Theocritus, of whoſe Doriciſms he had of courſe heard, That he was any otherwiſe ac- quainted with them I cannot but doubt, ſince the poem, which he has cited to illuſtrate the peculiarities of that dialect, has not a ſingle inſtance of it, being wholly compoſed in the common poetic language formed upon that of Homer and Heſiod. Even if it had, the word adduced would not be relevant, it being of a different claſs. or declenſion, ſuch as, in the /Eolic aud Doric dialects, has the nominative in -A and -A, and the genitive in -AFO, -AO, and -A; and in the Tonic and Attic, the nominative in -H, and the genitive in -EO, -EQ, and -OT ; whereas the words alluded to in the inſcriptions have the nominative in -HE through all the dialects, and the genitive in -EOZE, - Ox, or -Orz, according as local cuſtom had con- tratted or corrupted it ; but to omit the T would be to ſubvert all analogy of ſpeech (1).

The form of theſe inſcriptions is not leſs extraordinary than the ſub- ſtance of them, they being both ſigned by the public ſecretary, and au- thenticated by the public ſeal (2), upon which is engraved the name AA- KEAAIMON. That the public ſcribe or ſecretary ſhould ſigu a public re- cord or decree for putting up an inſcription is very natural ; but that he ſhould think it neceſſary to put his ſignature to the ſtone itſelf, ſeems wholly inconſiſtent with the manners of the Lacedæmonians, or, indeed, of any other people poſſeſſed of common-ſenſe : but, even if they had been guilty of ſuch an abſurdity, they would not have done it iu the form here employed ; for the officers of the public afſemblies in the States of Greece did nor authoriſe their ſignatures, by adding the mere title of an office to the name of him who bore it, but by uſing a verb which expreſſed, not only the office, but the actual exertion of it at the time of ſigning the re- cord on which it appeared. Thus, in the inſcription of Minerva Polias, we have NIKO®ANEE MAPA®ONIOE no ro EPPAMATETEEN, and, in the burleſque 1 imitation of the proceedings of a public aſſembly, acted

(1) Te 3s Teprrns A. e, ro Thy — Twy Nur big -O exewv. Theod, Gaz. lib, II. (2) See Pl. VI. and VII.



by the women in one of Ariſtophanes's comedies, the herald proclaims the decree of a council, in which TIMOKAET ENIEETATEL ATEIAA'

EFPAMMATETEN, EInE EQETPATH. Timoclea prefied, Lufilla atied as ſecretary, and Soſtrata made the motion, which had been voted (1),

The putting the public ſeal to a ſtone, in order to authenticate it, ſeems + ſill more abſurd than the having it ſubſcribed by the public ſecretary ; but nevertheleſs, M. Fourmont aſſures us, that he found every day at Lace- dzmon inſcriptions with the names of the kings and magiſtrates, and the ſeal affixed to them. The author of the Recherches, indeed, obſerving that no other traveller or Antiquary had been ſo fortunate as to find a ſingle inſtance of it in any other part of the world, ſuddenly tranſmutes the ſeal to a buckler, notwithſtanding the diſſimilarity of its form to thoſe publiſhed by Fourmont, and the impoſſibility of aſſigning any reaſon for its being introduced. It is rather wonderful that he did not turn it into a cart-wheel, to which it has ſome reſemblance, and which, being the em- blem of Fortune, might, by a little of his ingenuity, have been explained to ſignify the various fortunes of the war; which theſe inferiptions are ſuppoſed to commemorate becauſe they ſay nothing of it. Fourmont, without doubt, would have made them ſpeak very plainly of it, had not the ſame cauſe deterred him which prevented the appearance of the laws of Solon, namely, the conſciouſneſs of his own weaknefs, which, how- ever capable he might think it of forging titles and proper names (and ſuch are all the inſcriptions publiſhed), ſhrunk from the encounter of gramma- tical accuracy, to which any n like compoſition would have en him.

The next monument to be conſidered is another of the votive ſhields abovementioned, which is inſcribed with the name of Anaxidamus, the ſon of Zeuxidamus, who reigned at Sparta towards the cloſe of the eighth century before the Chriſtian æra (2). Underneath, upon the (baſe, ts his

(1) Theſmophor. Vſ. 372. The reader may obferve in theſe two quotations the nice employment of the tenſes. The verb in the inſcription, being merely to commemorate, is in the Aoriſt ; but in the decree, it being to authenticate, it is in the paſt imperfect. The. ation, when commemorated, was completely paſt ; but when empleyed to authenticate, fill exi Ws though pat with reference to the promulgation of the as authenticattd.

(2) See Pl. VIII.



pedigree : which, differing entirely from that given by Meurſius from Pau- ſanias, has affarded matter of much triumph, to the defender of theſe in- ſcriptions. The difference, however, proceeds merely from a blunder of Fourmont, who, caſting his eye careleſſly over the prolix pages of Meur- ſius, and obſerving the name of Anaxidamus, the ſon of Zeuxidamus, to follow thoſe of Eurycrates and Anaxander in the catalogue of the Agidæ, and not attending to the words ex altera familid, confounded the two royal houſes that reigned together, and transformed the partners of Archidamus into anceſtors, Hence the pedigree is, Archidamus, the ſon of Zruxidanus, the ſon of Anazander, the fon Eurycrates ; Whereas, according to all an- tient authors who have ſpoken of theſe princes, Zeuxidamas and Auaxi- damus were the ſon and grandſon of Theopompus, who were of the houſe of the Proclidæ; and Eurycrates and Anaxander, their contemporaries, the fon and grandſon of Polydorus, who was of the Agidæ (1).

The peculiarities in the word EYPIKPATEO have becu already noticed, and, I believe, are wholly unjuſtified by anticnt authority. That of the A for the Z in AETKEIAAMO is, however, authoriſed by the Zanclean medals, from which Fourmont undoubtedly took. it, He did not, how- ever, recolle& that the Zancleans were an Ionian colony, whoſe dialect fa- voured the clifion of the T; whereas the Dorians would certainly, in that age, have prefixed it to the A, as they always did to expreſs the Z of the other wrecks The Z too in this name muſt have been compoſed of the F and Z, an Fe K and E, as it is derived from ZET La, the regular future of the verb Zk TFH a, otherwiſe written ZEKTTNTMI.

The foxes and ſerpent, repreſented on this ſhield, allude to a ſilly fic- tion, probably the invention of later times, concerning a public ſacrifice, at which theſe animals appeared mitaculouſly upon the reſpective altars of the Meſſeniaus and Lacedæmouiaus, to prognoſticate the eveut of the war in which they were engaged (2), ; *

The laſt of theſe inſcriptions is alſo upon a votive buckler and its baſe, which contain the name of the city Lacedzmon, and of its king Archida- mus, the ſou of the great Ageſilaus (3). This prince was killed near Ta- rentum in the hundred and ſixth Olympiad (4), about three hundred and

(1) See Meurſ. Reg. Lacon. c. XI. and XVII. (2) Sce Apoll. lib. II. ſ. v. fifty-

(3) Sce Pl. IX. (4) Pauſan, lib, III. p. 230.



fifty-fiveyears before the Chriſtian æra; ſo that he flouriſhed when arts and letters were in the higheſt ſtate of perfection: yet the form, both of the ſhield and the letters, is as rude and barbarous as any of the others. The ouly eſſential peculiarity, however, conſiſts in the OO for the Q in the word AAKEAAIMOON, which, I believe, cannot be juſtified either oy authority or etymology in words of this claſs.

I ſhall now truſt to the candour of the reader to decide whether or not I have judged right in rejecting the authority of theſe inſcriptions. When I look them over, I am inclined to think that I have ſaid more than enough to detect them; but when 1 conſider the pertinacious obſtinacy with which forgeries, equally bungling, have been defended againſt perſons of ſo much greater learning and ability (1), and the. daring confidence with which others, long ſince detected and exploded, bave lately been adduced as au- thentic compoſitions of remote antiquity, to ſupport the wild paradoxes of viſionary theoriſts (2), I am. apprehenſive that I have ſaid too little.

It has been my endeavour to avoid avy inſulting reflexious upon the con- duct of thoſe learned perſaus who have quoted theſe inſcriptions as authen- tic ſources of importaut information; for, though it is the duty of every impartial inveſtigator of truth to expoſe fraud and detect error whereever he can find it, yet if he can accompliſh his end without wounding the feelings of any man, or the reputation of auy writer, his merit will be the greater. Fraud, indeed, deſerves no favour, being little leſs criminal when gratifying vanity thau when-gratifying avarice (3) ] but of this Jam inclined to acquit every one, in the preſent inftance, except. the original author, Fourmont, whaſe waut of genius and ability will, I fatter myſelf, never be Ban urged as. a prope of his ſincerity; for that which excites our

( 1) See the controverſies concerning Phalaris, &c.

(2) See Divine Legation of Moſes demonſtrated, and new Syſtem of antient Mytholog y; parti», cularly the latter, vol. II. p. 229, and vol. III. p. 77, of the firſt edition. The former does indeed tell us, that bis Letter of Alexander has been ſuſpe&ed, and offers the beſt ar- guments that he could find in its defence: but the latter boldly quotes the bungling and long-exploded forgeries of the ſecond century of the Chriſtian era as the certain and unſuſ- \ pedted compoſitions of remote antiquity ; though he, as well as every other ſcholar, muſt have known that they were fraudulent, and could only hope to avoid detection by the ob⸗ ſcurity into which they are deſervedly fallen. {

(3) This conſideration will excuſe the preceeding references

5 admiration


admiration at his forgeries, is not the ability employed in compoſing them, but the impudence exerted in publiſhing them; and this is a quality which. generally prevails in an inverſe ratio with the others,

That the authors of the Feune Anacharfis, and the Recherches far les Arts, (hould, by being the dupes of the impoſture, become partakers in, it, is extremely to be regretted, as both theſe learned writers have rendered conſiderable ſervices to polite literature, and are, I believe, both alike in- capable of any intentional guilt of this kind.

The former is a perſon of a very elegant mind, and has produced the only work extant upon the ſubject of antiquities that can boaſt of any ac- \quaintance with the Graces, The ſcholar and philofopher may indeed be diſſatisfied with many parts of his work; but the miſceHaneous reader will be every where amuſed with variety, and ſoothed with urbanity, without having bis underſtanding too much fatigued with deep reſearches, ' or trained by long and complicated deductions. He will find himſelf led. gradually over the wide and variegated furface of Grecian literature; but as his guide never analyſed the ſoil, nor examined the productions, he can only ſhow him general forms, and teach him unconnected facts; the firft of which intereſt but little, unleſs we know their mutual bearings and- particular relations to each other; and the ſecond not at all, unleſs we know the ſprings which gave riſe to them, and the ends to which they are directed.

The author of the Recherches dived deep into the matter, which he pro- feſſedly undertook to diſcuſs; and, had he confined his enquiries to that, he would have done honour to himſelf and ſervice to the publick; for many of his explanations of the monuments of antient art ſhow a degree of acuteneſs and ſagacity almoſt unparalleled. But when he invades the- province of grammarians, and endeavours to explain antient words, he al- moſt makes us doubt whether or not he continued to poſleſs the ſame fa- culties, ſo totally is he changed by changing his ſubject.


mall c T 1 ON VIL


ON THE GREEK 4172432

HE Decree of the Lacedzmonians againſt Timotheus, as it is pre- ſerved in Glareanus's edition of Boethius de *. 155 I. c. I.), is,

with marginal variations, as follows:

Emre 0% T :uodeop M Wapayiuevop 4 ray optrepy - Ay ray TANGIAY frOAT1V GTHLUTEG: Y rav di MAY dn xe xbagigu, wamogprpopurrop Worupuniev urayws?, fc i x T Tag cxxocip ra vewv de Te Top Touxopdap, 3 reg uu rg Ter WY ul αο⏑οννe, Y vom ayTh amhoayy & TETELY (46) ,das pOATY £0 xpupee]op Tuveigews! TETE pirtce bictgagw., Avri Y Erogpuovice Woe diſigęi po apoicay, IIa p 0s tv Tov ayave

Tap EMeuowicp Azupe]eop" 217 x08" dutnpigalo Tay TH poly xidyows , Tas yap Eepira cduvev u erlexarop Niop didaxnv i Era

Wepi rr Tov Caoinigy N Tov prilopep* e Tyobeev. Ema-

ve(05]ou Tt rd tina A txravup rap* Whpiog ap 's IR

he T OV e xagdoy ace. To yap r Hagog Jen Te.

reha. 66 re Trafſus e: : T. 441 X&ANWY „rc, Li. ä

ro e radial og & eyopwy”',

Oy * )


x e ro

cuyigal fuer "Aapaog

7.3 , ap.

robs Baring

rd pyTopay T "yep © TigiT]ap

  • ulav Papug




Gronovius firſt endeavoured ſeriouſly to reſtore It, and, in the Preface to the fifth volume of his Collection of Greek e publiſhed i it as


Emre: T. ο˙ MAD, i e ir Tay a lie- wie TAY was pay

a ru α OS , % Tar did Tav arg xopò dv rh @TogpePojuerogy vob e DT] Aupodvelcu rap axocp Twp „lp, dar rip Wohuxopoicp % Trp Kouveralop Tw pb- Atop ayer x woxihav avri omMoup N vera appiiniſai THY “, i xenupar Top TUVIOTApevop THY Ta case Race ori Top evappovie Wor Trav d paid WagoxAnbeig Ge % ir Tov d yave Tap Encuowins Adporgop dE Inge


9 3


mivod]o rd rd ub dagen, Top Tag Eepinap winap du by du rug vip di- daxxe* di don rau p 1 ri r Twp Haανν, % Twp ige Hi b Ti- Þrov, 8 oe ating JN J rar ddr ,˖mç᷑ inrapsi rap apo Cr 0196 paves T#p zur. on ira 10 Tap won dgop tpaiv haet, ir rd Emaproy imer f To Tay u h ib, hi rer ragd vlnr KNiop a ywruye

In the year 1977 a more correct copy was publiſhed from ſome Manu- ſcripts at Oxford, accompanied with variations found in other Manuſcripts belonging to that Univerſiry ; and a critical and explanatory Commentary”

by the learned and reſpectable Prelate who publifhed it. This copy, with the variations, was as follows :

Emudy' o Tinobrop o Mianoop magyyp og Hi ah, Wow Txy πτν — a wouy wipars xa: Tus

T Eris . « Ov Codd, al, Oxon.

  • mafarure Cod, Bal, 3 ow Tar

Codd. al. aua, Codd. al.

d. ry ſaxogòar A a ogpepoptrrop π Poriew® tg Nuance? rag avong Tor ror" d re rap worgopting nou Tap navoruraup* To" pwekiop eyerre* nou πũ·ẽꝛ ra yr! a a TTHEYCp au . Tay To u d arr; Tag wagons worran"” evTiaTgoper* aporbay, mapa, bt na roy

eryover” Try Exuoiying Aupeargeop empire”? Fox

caro rar T0 pra Ramat Tev Tap Tee od un edu Top veop dd f , As, b win? TouTow Top NνE,¾ p u reg pop fag .! T pobeov emavaxarai” d xa: tay eliza eq u dH her Tag Wepirap uvnronmopumeo? rag %% o ane To rag Wop Gapog oper euhabira:” ela Znaprav s, 11 T9 ps" xako verav jpuerroſe”* ragage r NN ayovroy"',

34 un Magd. & cater, 5 Bal, 8 al. 37 js al, areros al.

2? Bod!, h Magd. Bal.

ray fe 191 xxo r v EG

s pway Codd, al. rare Cod. Bal. eber Seld,, 7 aw owner Cod, Magd. may Magd. - car al, 9 Avuayuitar Seld. % „ Magd. 1 gorvxoptag Seld. 2 ve-. roraray Magd. 3 % Magd. as Toys al. „ mana» Magd. “ a- gun. Seld. apframmures Bodl..

  1. 7 ygoudre; Seld. xuparo Bod), &

Magd. „ enn Magd. wo- ar Bal, ora, Magd. . 4 Bodl, © arrzo0gores Bodl, arrige- e Dal. aworgoPor Seld. * wa- au; Seld. ane , Magd.

  • £7064 Magd. * Seld. & Bal.

arri cæter. Magd. Nera cæter. *5 g Bodl, % al.

  • 7 Seld, A Carat cæter. -

vara ta Bal. &c, CCC. © drum wwpns CCC, „er al, 30 wTop Bodl. & Seld. 3* en Bodl,

  • αντνο Seld. Seid. nige al.

3* Seld, Tapagipas. BOdl. ragapras

To this the learned Editor, * order to give his reader a more complete. and accurate idea of it, ſubjoined the following re formed copy, reſtored to-

what he thought its original orthography : : 7





This; however, only ſerves to prove/that the learned Pralekn dd mou oh.

actly know the value of his own publication; for moſt of his emendations

are cither unneceſſary, or tend to the ſame end as thoſe of the old trans ſcribers, that is, to eject every curious provincial peculiarity, not readily

underſtood, and to fill its place with a word from the more known dia-

lets. Like other editors, both antient and modern, * found it more on to alter than explam. -

The change of the © to tha * is unneceſſary ; 3 Rs though the Won i monians pronounced theſe two dental aſpirates in the ſame manner, it does

not appear, from any genuine monuments.of their writing, that they con-

founded them in orthography, or expreſſed them by one ſign, any more than we do the r and sn in the words FACTION and FASHION. The ſame may be ſaid of the change of the 1 for the T in all che in-

ſtances where this laſt vowel is uſually employed; for Euſtathius tells us,

that it was the practice, in the later Doric and folic, to put the I for the T(1);

(9 0¹ bs preyraper reife xaT To duppos doe, bee, ces, .es e n A%) 99% T0AVTH daf rica, KAb RON As P · 1913. J. 32. 3 all

1 1 9

” t


and the uniformity of it I copy of the Decree ſhews that it was in- tentional,

The inſerting the common aſpirate too, and not the "Rr is im- proper; for both theſe letters were dropt from the alphabet nearly at the ſame time, and neither of them occur in inſetiptions of ſo late a date as this Decree, unleſs indeed it be upon ſome coins of Elis, Heraclea, and Tarentum, the age of which cannot be aſcertained, and the columns of Herodes Atticus, written in imitation of the antient orthography (1). It was alſo cuſtomary in the antient dialects to drop the aſpirate from the conſonant, as has been ſhewn in the inſtances of the Zanclèan and Theban medals; whence I have no doubt but that MIT OE, which occurs (in the. genitive caſe) in the manuſcripts of the Decree, for Mreoz, is the true word, and not MYLOE, which the Editor would ne though it has a different aud incompatible meaning.

The change of the T to the A in MOIKITAN is right, and alſo that of the A to the O in the Jaſt ſyllable of KANOTATAP ; but the ſubſtituting an E for the A in the firſt is wrong, Gronovius's reading KAINOTATOP is probably right.

ATIMAZAE ſeems to be the proper form, and not ATIMAZAET, the ſenſe requiring a paſt imperfect rather than a preſent tenſe, and the omiſ- ſion of the augment being common to Homer, Heſiod, and Herodotus,

Kd, or KIOAPIKEIN, is alſo more conſiſtent with the roughneſs of this dialect, and more conformable to the antient terminations of the verbs in -KQ (whence the future - FC or -KEN) than KIZAPTTIN, given by the Editor, or KITAPITIN, which one manuſcript has, and which is leſs ob- jectionable.

AlAlPpEEIN for AAT KEIN, or ATAEKEIAN, is too „bene an alteration, if any alteration were neceſſary. which none is ; for the latter word is juſ- tied by the authority of a Manuſcript, and accords perfectly with the context. Even the firſt may poſſibly be right; for, though I have not met with ſuch a form as AIAEKEIE or AIAEKEFIE, the termination of theſe abſtract ſubſtantives in -x is as conſiſtent with the idiom and my of the language as in +A, 5

(1) 'To theſe, perhaps, may be added the Heraclzan tables, which have both aſpirates ;

but the age of them is uncertain, TIOITAN



MOITAN the Editor has turned to nNOION. or , The old Baſil edi- tion of Boethius has NOLAN, which may be the Doric contracted form of HOTANN, the fame as NOIEQNN, and therefore right. I prefer, however, the reading of the manuſcripts NOTTAN,. confidered as the accuſative ſemi-

nine of the participle. aoriſt, contracted, after the Doric manner, from

Holz AZ AN to Holz AN; and, by the change of the Z to the T, IIOITAN. A paſt tenſe is more ſuitable to this place than, a preſent; and it may refer to the preceeding ſubſtantive ATAEKEIN, or, ATAEKEIAN.

The change of NAPAKAE®QEIE to-IIAPAKAAQEIP may be right, as far as ſubſtituting the A for the E; but terminating, words of this claſs in P is unjuſtified by authority, and inconſiſtent with analogy, and. certainly not admiſſible in any dialect. 1 The ſyllable GA or @AP, which the Editor 10 600 as ufeleſs and i inex- plicable, relates either to the ſenate who enacted, or the ſenator who moved, the Decree; probably the latter; for the decrees, or qαπτ]ᷓ.Qdura, of the Greek Republicks, were recorded in the form of minutes, and had the mover's name adjoined to each, even ae they were voted, as, Inmoxgaras re IIAcr,,ẽ¶& two, &c. (1), 0 AlAAKKE in the manuſcripts is ok, as before Great ; the Editor's alteration to EalAAK ZE being the ſame as a change of 2yxs or duxe, in Homer and Heſiod, would be to. EHK E and EAQKYZE.

MEMYAZEQAI and EHANAPKAZAL, given by the Editor, are likewiſe wrong, the forms MEMYATTAI and ENANAKATAI in the manuſcripts. being more couſiſtent with the dialect, which. transformed the Z into a T, as well as. dropt the aſpirate. If any alteration is neceſlary in the laſt word, it muſt be merely the inſertion of the N—EHANANKATAI—accord- ing to the mode of ſpelling obſerved in moſt antient inſcriptions. I be- lieve, however, that no alteration is neceſſary; for, though this verb does not occur elſewhere, in the ſame form, we have other words of the ſame extraction and fignification, as axes, care, and avaxug, carefully j which, as

Euſtathius obſerves, are from the ſame root as ayaZ and arg,, words: which do not imply, in Homer, the office and power of a king in the

(>) Demoſth. in Neær.





preſent ſenſe, but merely-a,curator, or fuperintendant (1). The future in E or- KE proves thit'the verb avecvw was, at fome period, or in ſome dialects, terminated in -K, and, by the variations common in the Greek tongue, in -KER and-KAHH; ſo that EMANAKATAI was probably the regular Aoriſt inf nitive, in the Laconian dialect, of the verb which 6 igni-

fied that exertion of authority by which the Kang * ephori were to compel 

Timotheus to quit the city. eee

NETON in the manuſtripts is only wrong in the firſt letter, which ſhould be a B, TON, ot (as in the Etymologicum magnum) BETTON, the regular Laconian form of FEBON (2). Probably it is ſo in the manu- ſcripts, for the barbarous N and B of the lower ages are eafily miſtaken for each other. ELON, ſubſtituted by the Editor, is taken from a note upon Heſychius, who gives BEEON as Laconian for EQOE, and AON for naeox, by Which he ſeems to expteſs rather the vicious pronunciation, than the eſtabſſhed orthographyyof that people, -

TAPAPETAT in the mauuſoripts is right, and not TAPATTETAT, given by the Editor, it being the Laconian form of the ſecond Aoriſt ſubjunctive middle, and not the preſent of the ſubjunctive paſſive. In common Greek it would be rapzoyrar, or raparyres, from rapacow, or Taparrw. This ac-

cords with the precerding verb ETAABETAI, or enaCyra: Though 

theſe forms are called ſecond Aoriſts, they have almoſt always a future ſig-

naiſiestion in the early writers, as in wy 4v- pry Ar , tyomra, Herodot.

lib. I. c. 3. "Oprvoros yo priiynnccs e dae £740 W vo HD. rug av h Lov Yyren. Ibid, o. 29˙ TY

(1) Availa; tn Thr; rom 64 Trart4ct beato avaru;, nity Adder xl Twy rr. Euſtath. p. 21. 15. See alſo 1425. 56. N o crew vn aoghueur ([lipravdpor), He- rodot. lib. I. c. XXIII.

(2) BEETON 20 Lario dre Aae & 3. BET TON. Ale. I have before obſerved . wwe double Powe ; this word, ſimilar to that of 1 in our own language.



— —ä—8— —





» Fig- 2.


big .








4 8 , F 4 2 2 * l * - *

  • * *

. 6.

  • 2 *

. % o

  • . > * q * :

1 ”

  • .

7 „ ** & Wo P.

  • 1 A
  • 4 *

$ * *

  • 4

« 0

  • fl * PR 1

by J 1 „ * . 0 F 4 * *

, d, LE and . 9 4

1 1 *

  • * 4

. * % o *, - by n 7 a

W 4

— 5 —




  J at 4  

« 6 TEPREARTE £ +.

_ Beg - 7.




\V VY AAATAM---- AZVA AZ... S ia 117 IShov ALF -:OEE AAo NV)




— 1


- 4 & EI ee —— ——

EY =







5 N een TEEp S e eee S PN &




Ak Þ



SY ee eee M A \ AN ANA AOA K






gov V N

—__— —— ">

— 8 SY XS RN





= e pats eee e boy, Ac ic TAPXOCAPICT 9 -+ \ AMP/OCAAMXOY



Oe D “


Ixc(opſoca AA MVH K — Ihe


= 2A AY KeoPrOCKAEOA oY = eee ef


OoPol_ > 9 5 & AMOPIA OC, AMoGy oy

xt opſocx At NN


mus 1 —

PPP ˙·˙—˙⁰ ̃ oO —ů———ů

r ˙ ¹w] XA .d. GEE — ß——³1 nn I II nr

















K &.2a AMo(APICTANþPof i

A ee



y CME NooPol


2 ——_—_

18 - ( —— = _> x N- r

Aõοα ?


RO — — — —





1 Wl

dae ee OY "OP BE


MAPAPo] Free




Ss # +

PL. vn.


ANM*P[Loha Mora KoY CONNY po

006 IK &y roy


P04 of Roy


PY5 ol



Do p 1440GAL&ANEE0



FAMAP jon DNνẽð






TAN HN fToraxef



(To ννον


RMogY Fo



IAN eee

XA /


LY Ko ra xohe YkeyrIY


TNA pioL FN Roy IKE ILT o’



AAM929[L05xR Ad ADP













[ ce


  • *


— — — 2 — — %

. , 0 — 4 2 ** — 3 — — „ — * . - », „ ,

  • ";

„ -

  • *


  • %

+4 Y


  • 3 4

0 - 2 — PE

fi : *
  • * ,

— " 4 Fg

  • 2 *

1 — - * * 1+

  • bo

% 1 o

  • .

« P = o F- = - . « 

. % „

W = 0 . - - 2 .

ye ; 2

% 2 „% i i — — —— — F -— _ - — — — On us. 2 2 3 ¶ ꝙ ——— 5

A + - 2 - ——— T—_ — — — — im. ů *˙ . * - —_— — ů 1 5 " y

  • — 5 hy

— — - a © -

  • .

—— x ——-„— —— —„—

4 yp - 7

\ a

138 „



AM Bkp14 AMOY PEYR 50% A


PL- X.

Front matter

-es, ditto of ditto participles in 52 Aſpirates *® Ft 77 =» 8 Aſpirates, metrical power of 29, 51



" Page    Page 1 

A 1 - 19 Aſpirates, where elided 40, 68 A. Alpha "os 17 Aſpirates, where loſt - -a, terminations of nominatives Aſpirates, doubled - _ ſingular of nouns maſc. in 39 B - = - -a, ditto of genitives ſingular of Batrochomyomachia - ditto in - — 40 B or * WM - -a, ditto of nominatives ſingular C, whence defived _ =- of nouns feminine in 43 C, how pronounced = A, privative _,-., - 65 Cadmeian Letters - A, collective, or afpoigmov 92 Czſura © - Ain, Phoenician — 17 CH, French = - Alcman, the poet - 36 Conſonants, double 17, 18 Aleph, Phoenician - 17 Conſonants, metrical power of 26 Ambiguities - - $895 Cratylus of Plato . Amyclzan Chronicle - 116 D - - - 4, 8 ar, terminations of genitives Datives plural - 32 4 plural of nouns in 40 Decrees; Greek — 135 -a, ditto of ditto ſingular of Diale&, Ionic . = 38, 47, 55 ditto in 5 - 39 Dialect, Doric — 38, 40, 54 -&wv, ditto of ditto plural of ditto Diale&, Æolic 12, 38, 63 feminine in $ 43 Dialect, Attic 38, 47, 55 Arcadian Letters - 120 Diale&, Hellenic Is 47 Articles L 79 Dialeas of Homer - 7 Articulation - * 3 Diphthongs - 18, 41, 70 Arundelian Chronicle - 119 axzorz, «= 0 32 - ag, terminations of nominatives . - - 19 ſingular of nouns maſc. in 39. E, French - - 17

-u, terminations of nominatives ſingular of adjectives femi- nine in — % +4 2



79 as a W. 793.

1 7124 ;

iv I N D


uc, terminations of nominatives ſing. of participles maſc. in 52

e, ditto of ditto of ditto ſe-

minine in 93 54

-i, ditto of datives plural of ditto in 7 (

«60, ditto of genitives ſingular of nouns maſculine in 39

-10, ditto of the ſecond perſon ſing. pafl. of verbs in 45

-tw, ditto of genitives fingular of nouns maſculine in Dy tog, ditto of ditto of ditto in 44 ic, ditto of nominatives of ditto in 39

-% Or EFL, ditto of ditto of ditto in - Extenſions, arbitrary - 44 » /Lolic Digamma 10, 11, 12, 39

F, ditto, metrical power of 35- I, Roman - 3. 11, 33 Future teuſe - 107 G, conſonant » 4 P, aſpirate - 17, 69, 82 GH - - - 10 Grammar, Greek - 105 I'S or Z - - IO, 32 11, Phoenician 899983 8 H, Etruſcan — - 8 H, Ronan. . 7 H, Greek aſpirate - 8, 12 H, ditto, repeated - 43 H, ditto, reduced to a mark 9 II, ditto, arbitrarily affixed = 41 UI, Gieek vowel 998 12

38, 88


Page -y, termination of nominatives

ſing. of nouns feminine in 118

ic, ditto of ditto of nouns maſ-

culine in - 39 -y5, ditto of ditto of adjeQives ditto in - . aa: 96

Heſiod amended 25, 81, 103

Homer amended 31, 33, 39, 41, 46, 56, 57, 60, 62, 67, 68, 73, 75, 79, 84, 87, 91, 1

Homer's text, preſent ſtate of 30, 77

Hymns of Homer - 30, 60 5 — - #5 20 I, conſonant - - 20 I, ſubſidiary — - 75 I, changed to T - 3 ; I, iota ſubſcriptum * +» 26 IsT&KpAntt; . - 115

-, terminations of the accuſa-

tives ſingular of nouns in 54 Infinitive mood -

Inſcriptions of Herodas Atticus 96

Inſcriptions of Fourmout 111 Inſcription of Minerva Polios 126 Inſcription of Scipio Barbatus 7

107 Inſcription, Delian - N - Inſcription, Duilian - 4. 7

_ Inſcription, Eugubian 5, 120 Inſcription, Heraclean 8, 29, 43


Inſcription, Sigean 10, 92 Inſcription, Veletrian 4, 16, 58 Inſcription, Venetian 10, 16.

og, terminations of adjectives in 70 -15, ditto of nouns in - 53


44 ——

5 bs. 96. 78



-19g,. terminations of adjectives in 76 K e 1 4.5 KH . » .2> ol 8 KN „ - 30 K ͤ or 3 - + 10, 33 „ 16 Lacedæmonian decree 62. 135 Lacedæmonian magiſtrates 124 Lacedæmonian manners 114 Lacedzmonian ofthography 15, 124, 136

Leſbians — eie Letters, adſcititious 65 Letters, Latin _ 121 Letters, ſquare and round 5 Licence, poetical - 59 Liquids - - I5 Liquids, metrical power of 29 M - 23 16 Margites of Homer 8 Medals of Achæa - 13 Medals of Agathocles 48, 105 Medals of Alexander 11, 40 Medals of Alexandria Troas 58 Medals of Amyntas - 40 Medals of Antinous- - 19 Medals of Capua - 35 Medals of Corinth - 5 Medals of Cos - - = 17 Medals of Croto - - [7 + Medals of Dionyſus +4446 Medals of Elis - 12, 76 Medals of the FAAEIOI 12, 76

- Medals of Gelo - 46

Medals of Gnofſus + 906 Medals of Gortyna 114

Medals of Heraclea in Italy 8 Medals of Heraclea in Sicily 20 Medals of Hiſtickaa 33 Medals of Leſbos 5, 8, 18, 96 Medals of Lyttus hee Medals of Naxus - 10 Medals of Oaxus II Medals of Perdiccas 40 Medals of Philip oy 46 Medals of Poſidania 16, 114 Medals of Rhegium += 17 Medals of Side eln Medals of Sybaris - 16 Medals of Syracuſe - 5 Medals, Syro-macedonic 20 Medals of Tarentum . Medals of Thebes — 40, 86 Medals of Zanclè 8, 10, 33 Meflenian war - 123 Metatheſis 44, 66, 72, 82 Mette, laws of - 23 Middle voice - HU BQF- - - Minos - e013 Mithridates - 20, 27 Mutes - - 8 N - - Reg 16 N, paragogic 53, 58 Nouns, flexions of — 38 QO - e412::; 3 O- neya, or 2 12 5 19

-o, terminations of genitives ſing, of nouns maſculine ian 43

Aue by 111. V "64 -

1 I

5 Page , terminations of nominatives

ſing, of nouns feminine in 43, 53 -w, ditto of accuſatives of ditto in 50 bee, ditto of genitives of ditto

maſculine in - 43 - wv, ditto of ditto plural of ditto in 40 -wys ditto of nominatives of ad-

jectives comparative in 45

- ditto of ditto of participles in 53

or, ditto of the third perſon plural of verbs in $3» 55 Orthography, Greek 14, 29 Orthography of Homer 77 Orthography, Engliſh - 14 Orthography, French - 14 Orthography, Italian © +» I4 Orthography, Spaniſh - 14

-0;, terminations of nominatives ſing. of nouns maſculine in 43

oc, ditto of ditto of ditto neuter in 43 -o, ditto of genitives ſingular of

ditto maſculine in - 44 -ov, ditto of ditto of ditto in 43, 46 -o, ditto of the ſecond perſon ſingular paſſive of verbs in 45 - obs, ditto of genitives ſingular of nouns maſculine in 44 =evg, ditto of ditto of ditto fem. in 5 3 -ovg, ditto of nominatives plur. in 45 -ovg, ditto of accuſatives plural in 46 -ovg, ditto of nominatives ſingular of participles in -


oba, ditto of ditto of ditto fe-


minine in — 


X Page o terminations of datives plural in - 54

eue, ditto of the third perſon


P. A . fry A. An,

T. . 62 .

plural of verbs in 53, 55 P - 2 13 4» 6 Parian Chronicle 119 Particles, paragogic - 50 Particles changed 79 Pelaſgian Letters 120 Perfect tenſe - 107 PH, or © 7 1175 8, 13 Pleonaſms - - 44 IIN - - - 39 Prepoſitions, change - 79 Pronouns, flexions of - 48 Pronunciation of vowels 21 Pronunciation of Homer 77 NZ, or Y - — 32 nT - - - 30 K-68 - ee Records, Greek 7 126 8, Latin - — 11 8, ditto, metrical power of 32 D, Greek — - I4 E, ditto, metrical power of 32 E, elided 62, 63, 75 Scythians - - 14 SH, Enghiſh - 14 TK : : 2 33 LM - - + - Sophocles, amended - 103 2 2 ? "” - for T - 95 — 4, 8 Tenſes, Greek — 108



TH, or © — 13 TM = - 30 Tone — — 3. 16 TE, or Z - - 32 3 — - — 20 V, Latin „„ v0" V, Pelaſgian - - 121 V, Arabian - - 35 T- — "Wi 12

, Greek aſpirated 12, 120 Y-:Xov, elided 41 T ditto, ſubſidiary _ + 54, 75 I ditto, changed to I -. 499 Verbs, flexions of _ 106 ur, terminations of accuſatives

ſing, of nouns feminine in 46 


  • vii

Page Vowels - - 16 Vowels, long - LOS - Vowels, long by corruption 73 Vowels, Greek, pronuciation of 20 Vowels, metrical power of 24 rug, terminations of nominatives ſingular of nouns in 49

us, ditto of ditto of adjectives in 76 de, ditto of accuſatives plural of

nouns feminine in 46 W, Engliſh = 11, 121 W, Welſh - "5 ol X, Roman 5 10 x, Greek Chi - s 7 X, Greek Z - — 10

See also

Unless indicated otherwise, the text in this article is either based on Wikipedia article "An Analytical Essay on the Greek Alphabet" or another language Wikipedia page thereof used under the terms of the GNU Free Documentation License; or on research by Jahsonic and friends. See Art and Popular Culture's copyright notice.

Personal tools