-NRLF 
 
 6391 
 
.U•^ 
 
 Studies In Archaism 
 
 IN 
 
 Aulus Gellius 
 
 WALTER EUGENE FOSTER 
 
 y 
 
STUDIES IN ARCHAISM 
 
 IN 
 
 AULUS GELLIUS 
 
 By 
 
 WALTER EUGENE FOSTER 
 II 
 
 Submitted in Partial Fulfilment of the Requirements 
 
 for the Degree of Doctor of Philosophy 
 
 in the 
 
 Faculty of Philosophy 
 
 Columbia University 
 
 NEW YORK 
 
 I 9 I 2 
 

UXORI CARISSIMAE 
 ETHEL FOSTER 
 
 254402 
 
STUDIES IN ARCHAISM IN AULUS GELLIUS 
 INTRODUCTION 
 
 The following pages, which have to do primarily with 
 archaisms in the vocabulary and the syntax of the Noctes 
 Atticae of Aulus Gellius, are intended to be, in general, sup- 
 plementary to the work which has already been done in this 
 field by Professor Charles Knapp, of Columbia University.^ 
 Not only has Professor Knapp aided the author of this paper 
 with suggestions and encouragement, but he has, as well, most 
 generously turned over no inconsiderable amount of material 
 which he had gathered from time to time, with the intention, 
 at some later date, of publishing it and thus completing his 
 studies in the archaisms of Gellius. - 
 
 An entirely independent examination of the material has, 
 however, been made. In some instances, the results have been 
 merely corroborative, in some supplementary, and again in 
 others corrective. In the course of the investigation, again, 
 certain phases of the archaistic elements in Gellius which had 
 been either entirely overlooked or merely hinted at by previous 
 investigators receive due emphasis. 
 
 The Introduction to this paper (pages 1-28) deals with 
 certain general questions relating to the archaistic movement 
 in the second century A.D. and. in particular, with the varying 
 relations of Pronto, Gellius, and Apuleius to that movement. 
 In Part I (pages 29-47), which deals with archaisms of form 
 and vocabulary, the lists given are intended to be supple- 
 mentary to the lists published by Professor Knapp,^ or cor- 
 
 ' See the Bibliography below, page 65, s. v. Knapp. 
 
 ' It should be noted that it is impossible to study archaism in 
 Gellius without at the same time examining carefully Fronto and 
 Apuleius. For the relation of Gellius to Fronto see below, pages 16- 
 20; for Gellius and Apuleius see below, 21-23. 
 
 ' Drisler Studies, 141-146. For the citations in this paper see the 
 Bibliographical Appendix, pages 65-67 under the names of the authors 
 of the books and articles cited or quoted (some few books or articles, 
 cited or quoted but once or twice, have not been listed there; in all 
 such cases the book or article will be sufficiently identified in text or 
 
4 Archaism in Aldus Gellius 
 
 rective of those lists. Part II (pages 48-64) is devoted to an 
 attempt to set forth in full the archaic qualities of Gellius's 
 syntax, a task which no one else, so far as I am aware, has 
 essayed. 
 
 It is not the writer's purpose to attempt, in any compre- 
 hensive way, an explanation of the intense archaising fervor 
 which characterized the literary spirit of the second Chris- 
 tian century.* Nor, indeed, is such a comprehensive survey 
 any longer necessary. Nevertheless, some special remarks 
 will be offered a little later (pages 5-6). For the present a 
 very rapid resume of the history of Latin literature will be 
 made, to pave the way to what is said below (page 5) about 
 the reaction in the time of Gellius and Fronto in favor of the 
 Latin of the archaic period. First, we have the early, crude 
 efforts of the pioneers who are striking out for themselves 
 new paths in untrodden fields. The qualities of freshness and 
 vigor in part, at least, compensate for what may be lacking 
 in perfection of form and in workmanlike finish. Next comes 
 the classic, or the golden age. Now is the high tide of national 
 life. In this period, poetry and prose reach a perfection be- 
 yond which, except in rare instances, the genius of the race is 
 unable to go. Dignity of manner is combined with freshness 
 •^ and originality of theme. With the close of the period of ex- 
 pansion, there was a tendency towards the fixed and the con- 
 ventional in all spheres of life. This tendency is marked in 
 the realm of letters. 
 
 There is, therefore, in turn, ushered in now the age which, 
 taking the great classical masterpieces as models,^ pays less 
 and less attention to substance, and seeks more and more ex- 
 quisite refinement of manner and uniformity of method. In 
 the first Christian century, the opponents of antiquarianism 
 
 footnotes). The abbreviations employed in the citations will also be 
 found listed in their alphabetical places in the Bibliographical Appendix. 
 
 *For discussions of this matter, see e.g. Vogel, 19-22; Mackail, 233 
 ff. ; Teuffel, §345 ; Kretschmann, 2-4 ; Schmalz, 664-665 ; Stolz i. § 36 ; 
 Nettleship, Lectures and Essays (First Series), 279; Piechotta, 5 ff. ; 
 Brock, 25-35, 181-183; E. Norden, 344 ff., 361 ff. ; and especially Drisler 
 Studies, 126-141 (with literature there cited). 
 
 "Duff, 30. 
 
ArcJiaism in Aldus Gcllins 5 
 
 were supreme.*' Here the goal is symmetry, harmony, ele- 
 gance ; but after this refining process has been carried to the 
 extreme limit a reaction sets in. 
 
 In Rome the reaction against this super-refinement, under 
 the leadership of such men as Fronto and Gellius, contributed 
 no doubt to that reversion to the preclassical manner^ which is 
 the marked characteristic of the Latin writings of the second 
 Christian century. Cato, Ennius,^ and Plautus were studied 
 with diligence, and the more vigorous but simpler style be- 
 came, for a time, especially under Hadrian and his immediate 
 successors, the vogue among the literary classes. 
 
 It would be instructive and interesting if we could discover 
 in full the forces at work which produced men of the type 
 of Fronto, Gellius, and Apuleius, men who differed so widely 
 in their temperaments and literary gifts, and yet, in spite of 
 their marked differences, have a general likeness, in this one 
 particular: the archaic quahties of their style. 
 
 Some of these forces, probably — nay beyond doubt — the 
 most important, in the case of Gellius, at least, have been set 
 forth in the views of the writers named above, note 4, 
 and in the authorities cited by them.'-* These authorities, and 
 others, lay stress, and rightly, on conscious and deliberate imi- 
 tation of the earlier authors by the writers of the second 
 Christian century. 
 
 There is, however, another factor which, though noted by 
 some scholars,^*' has not received, it seems to me, the attention 
 which it deserves. I have in mind a kind of unconscious and 
 wholly natural archaism, which, I think, was inevitable to a 
 writer born and reared in Africa, as Apuleius surely was,^^ 
 
 * Xettleship, Lectures and Essays (Second Series), 74; Drisler 
 Studies, 137-138; Bernhardy, 327 ff. 
 
 ' Stolz, § 36; Nettleship, Essays in Latin Literature (First Series), 
 279 ff. ; Duff, 31. 
 "Vahlen, Ixxx ff. 
 
 * I do not regard Gellius as specifically an African writer ; see be- 
 low, note 51 ; also pages 14-16 in general. 
 
 "Compare e.g. Cooper, xxvii ff. In Piechotta, i-io, a good discus- 
 sion of the matter, the various authorities are cited and their views 
 considered. 
 
 "See e.g. Flor. § 18 (page 35, Helm). 
 
6 Archaism in Aulus Gellius 
 
 because of the conditions under which Latin became the speech 
 of Africa and remained in use there. 
 
 A priori there seems, to me at least, to be no valid reason 
 why the Latin language in Africa, in the second Christian 
 century, should not already have developed along lines such 
 as resulted in a speech, still Latin, to be sure, but yet distinct 
 in a variety of important particulars from the Latin language 
 as it had developed during the same period in Italy, and par- 
 ticularly at Rome.^- Analogies and comparisons rarely prove 
 anything, though they may shed a flood of light upon the 
 question at issue. 
 
 Ireland, for example, has been under English domination 
 for four centuries. Yet the language of Ireland to-day, though 
 officially English, varies through countless gradations from 
 pure English to Gaelic. Even in a country so small, so near 
 England, and so easy of access, so often visited by residents 
 of England, with so many of its own people visiting England, 
 the native language still leaves its impress, not only on the 
 oral speech, but on literature. For example, the numerous 
 ballads of Ireland^^ written even within the last century^* 
 furnish a rich field for the student of the early Gaelic lan- 
 guage and literature. 
 
 The Scotchman usually betrays his nationality in his speech. 
 Furthermore, the Gaelic has left its impress on the later 
 literary products ; for example, not only many of the poems of 
 Robert Burns, but comparatively recent works of literature 
 show the influence of the ballad.^'* 
 
 In Canada there is a rich store of ballad literature in a 
 patois which is neither French nor English but a curious com- 
 mingling of both.^"' To one who has travelled in Brittany, and 
 
 "Duff, 27, says: "If we pass to the fourth century A.D., it is not 
 for a moment to be thought that African Latin was identical with 
 that spoken on the Rhine or the Danube". 
 
 " See Hayes, The Ballads of Ireland, Introduction, 13-14. 
 
 "See The International Cyclopaedia, under Irish (Gaelic) Language 
 and Literature. 
 
 ^° Child's English and Scottish Popular Ballads is a great work on 
 the subject. See also Eyre-Todd, Scottish Ballad Poetry, 29-34. 
 
 "See Henry Drummond, The Habitant (New York, 1897). 
 
Archaism i)i .lulus Gellius 7 
 
 noted the dialect spoken there, which differs so markedly from 
 Parisian French, surely the claim that there was a marked 
 difference between African Latin and Roman Latin would 
 cause no surprise. 
 
 We need not go, however, far afield to find examples of 
 archaic words and usages in language. In out-of-the-way 
 parts of our own country, which lie outside the sweep of the 
 great currents of our modern life, survivals from earlier times 
 have for centuries maintained themselves almost unchanged. ^^ 
 Thus in the less frequented parts of Virginia are heard, to 
 this day, words which have been obsolete for two centuries or 
 more, but were current when the cavaliers of the Elizabethan 
 period settled in the New World. Many archaic words and 
 phrases may still be heard among the New England hills. 
 
 If, in spite of the railroad, the telegraph, the daily, weekly, 
 and monthly publications, such differences of language, dif- 
 ferences which, in some instances, amount practically to dif- 
 ferent dialects, can maintain themselves or develop in our own 
 country, there seems to be no a priori ground for asserting 
 that the Latin of the African province did not have marked 
 qualities of its own which differentiated it from Latin in other 
 parts of the empire, and, more particularly, from the Latin at 
 the capital. On the contrary the probabilities all point to such 
 a variation. ^^ 
 
 Africa became a Roman province in 146 B.C. From that 
 time on, the official language was Latin. ^^ The Roman soldiers 
 who brought Latin into Africa did not bring with them the 
 polished speech of the Scipionic circle. The rugged speech of 
 Plautus would more nearly correspond to the vernacular of 
 the Roman conquerors.-" The Latin which was first carried 
 to Carthage was thus still highly archaic."^ 
 
 In this province, separated from Rome by a sea voyage, if 
 
 "See Lodge (entire article): Cooper, xxviii, and note 3: Brock. 26, 
 note 3. 
 
 " Cooper, xxi-xxx ; Olcott, xv-xxi. 
 "Kubler, 161. 
 
 ** Cooper, xxvi f., and authorities cited there. 
 " Ott, Neue Jahr., lo;. 767 ; Cooper xxxv. 
 
8 Archaism in Aldus Gelliiis 
 
 we may disregard the Roman governors and their staffs, whose 
 influence upon the Latin language as spoken and written in 
 Africa must have been so sHght as to be practically negligible, 
 in addition to the legionaries, the peasants and the traders 
 constituted the principal Latin and Roman elements of the 
 population. It is evident, therefore, that, while the literary 
 language and the sermo cotidiamis of the cultivated classes at 
 Rome were relatively unimportant factors in the development 
 of the sermo Africus, the influence of the sermo plebeius was 
 very great. That the sermo Africus abounds in plebeian ele- 
 ments is, then, to be expected. To the cultivated classes at 
 Rome the sermo Africus would seem at once plebeian and 
 archaic. ^^ 
 
 At Rome, apparently, it was a recognized fact that the 
 people from different provinces of the empire had peculiarities 
 or tricks of speech which betrayed the place of their nativity.^' 
 Asinius Pollio's criticism of Livy's Patavinitas'^* is well known. 
 It is important to note in this connection that, in the opinion 
 of some scholars at least, this criticism was aimed, not at the 
 poetical coloring of Livy's style, but at his use of foreign 
 words. ^^ Lucilius in a similar manner had charged Vectius 
 with "Praenestinity".-® 
 
 "For the intimate relationship existing between the plebeian and 
 the archaic see below, pages 9-10. It is interesting in this connection 
 to recall the extent to which the language of Cicero's letters reproduces 
 the language of Plautus (see e.g. Tyrrell, Cicero in his Letters, Ixxii- 
 Ixxviii). If in Italy itself the sermo cotidianus of the cultured could 
 still in Cicero's time so strongly resemble the Latin of Plautus's day, 
 surely, keeping in mind Roman conservatism, we need find no diffi- 
 culty in believing that the Latin language as spoken in Africa for long 
 centuries retained evidences of its origin, as the plebeian speech of 
 soldiers, traders, and the like, of the second century B.C. 
 
 ^ Cicero, Pro Arch. 26, charges the poets from Cordova with a 
 foreign brogue; see also Spartianus (Vit. Sept. Sev. 19): sed Afrum 
 quiddam usque ad senectutem sonans. Jerome (Ep. 1.30.5) mentions 
 the stridor Punicus; Cyprian (Ep. 25) remarks: Latinitas et regioni- 
 bus mutatur et tempore. 
 
 "See Quint., Inst. Or. 1.5.56; 8.1.2-3. 
 
 ^See e.g. Duff, 639, note i. For this view Quint. 8.1.2-3 makes 
 strongly. 
 
 "Quint., Inst. Or. 1.5.56. See here Olcott, xvii-xviii. 
 
Archaism in Aiiliis Gcllius 9 
 
 On the foregoing pages (6-9) we have been dealing with a 
 priori considerations. That African Latin does in fact display 
 marked peculiarities has been shown at length by competent 
 scholars, after careful investigation, and full consideration of 
 the opposing view, long warmly urged in certain quarters. 
 References may be made here to the discussions of Cooper, 
 xxi-xlvi, passim; Olcott, xv-xxi ; and Kiibler, 161 ff. Cooper, 
 xlvi, sums up his elaborate discussion with a brief summary 
 of peculiarities in word-formation "which are now generally 
 recognized as characteristic of the African writers". With 
 his results Olcott, page xxi, and Kiibler, 161 ff., are in sub- 
 stantial accord. Cooper, xlv-xlvi, and Kiibler, 202, emphasize 
 the plebeian and the archaic elements.-' 
 
 The intimate relationship, which often approximates iden- 
 tity, that exists between the plebeian and the archaic Latin 
 requires no demonstration here.-® It is precisely this identity 
 that makes any study of archaism so difficult. Some have 
 been so impressed with the Plautine qualities in Apuleius that 
 they have not hesitated to assert and to attempt to prove that 
 Apuleius borrowed wholesale from Plautus words-^ and syn- 
 
 ^ In his edition of the Cupid and Psyche episode of the Metamor- 
 phoses of Apuleius (1910), Mr. L. C. Purser discusses elaborately the 
 style and language of Apuleius (Ixx-c). On page Ixxxiv he describes, 
 in vigorous terms, as wholly discredited the doctrine that there was a 
 special Latin style which might properly be labelled Africanism ; he 
 thinks this "phantom . . . may be considered as finally laid to rest by 
 the crushing chapter of E. Norden, Kunstprosa, pp. 588-598". As a 
 matter of fact, Norden does not deny (see 588) that there was an 
 African Latin in the sphere of language (in phonetics, use of words, 
 and syntax) ; he does, indeed, add that, for the most part, the views 
 current on these subjects seem to him "vorliiufig mehr oder weniger 
 problematisch", but he concludes with the words : "Doch das geht 
 mich hier nicht an : ich habe es mit denen zu thun, die von einem 
 afrikanischen Stil sprechen". Norden's discussion, there- 
 fore, does not cross the lines of the present paper at all. Brock, 164 
 flF., also is not pertinent to our discussion. 
 
 ^ For an admirable discussion of this important problem see Cooper, 
 xxi-xxx, with footnotes. See also Knapp, 151 (s.v. lac), 155-156 (s.v. 
 asf^er)iabilis), 157 (s.v. co)iiplusculus) , 162, etc. 
 
 " Desertine, 3. 
 
10 Archaism in Aldus Gcllius 
 
 tactical usages.^" But while Desertine confines himself tc 
 illustrating by examples the many striking echoes, resemb- 
 lances, and identities between Plautine and Apuleian vocabu- 
 lary and syntax, he is well aware that Plautus is only one of 
 many of the older writers^^ of whose works Apuleius con- 
 tains so many reminiscences. 
 
 It is important to note that the Metamorphoses are most 
 strikingly plebeian (and archaic) in quality. The fact that 
 certain of Apuleius's writings, the Apologia and philosophical 
 works, for example, are written in a style which much more 
 nearly approximates the classic norm,^- has been used as an 
 argument to support the theory that Apuleius's archaisms in 
 the Metamorphoses are entirely the results of intentional imi- 
 tation of the ancients, by the use of which he aims to give to 
 this work an atmosphere of strangeness, and perhaps of 
 mystery. If we could be sure of the chronological order of 
 the composition of the different works, we could discuss the 
 matter more satisfactorily.^^ Admitting the difficulties in the 
 way of definite knowledge as to the time of composition, we 
 may still hold that both the subject matter and the style of the 
 Metamorphoses seem more likely to be from the hand of an 
 author in the early flush of youth. The philosophical works 
 appear to belong rather to his maturer years. Whether the 
 Metamorphoses were written before his other extant works 
 or not (they may well have come from his brain when it was 
 riotous with the vivid imaginings of youth), one is on con- 
 servative ground who maintains that in writing these tales, a 
 work not intended for small coteries of scholars, Apuleius 
 allowed himself greater freedom, giving full rein to his fancy, 
 using the words and phrases with which his fertile mind 
 teemed, and not circumscribing himself by classic canons. 
 A popular book, for general circulation among all classes, 
 
 ^*Leky, 5 ff. For a more detailed consideration of Leky's views, see 
 below, pages 23-26. 
 
 " See Desertine, 5 f. ; Helm's Praefatio to Flor., xxii ff. 
 
 ^ Brock, 185, note i. 
 
 ^ Helm, in Praefatio to the Florida, x f., discusses the matter ; see 
 also Teuffel, § 367.1 ; Purser, xv-xxi, evidently regards the Metamor- 
 phoses as an early work. 
 
Archaism in Aiihis Gcllius ii 
 
 recounting the supposedly contemporaneous adventures of the 
 hero, who often is strangely confused with the author him- 
 self, would hardly have been written in a style which, because 
 of hundreds of words really rare and obsolete, would make 
 every page seem not only unfamiliar but often almost unin- 
 telligible. To me, at least, the more natural view of the mat- 
 ter is that in the Metamorphoses we have, on the whole, the 
 language and style natural to Apuleius, the resultant of birth, 
 training, and his own natural literary gifts. ^* Of course, we 
 must not forget that his training included a wide reading in 
 the Latin classics. His quotations from the old masters of 
 Latin and his allusions to them prove how intimate was his 
 knowledge of their works. ^^ 
 
 Miss Brock,^^ however, takes a position almost diametrically 
 opposed to the views outlined above. Many of her assertions 
 seem to lack verification and consistency. She sometimes 
 brushes aside as trivial and hardly worthy of her serious con- 
 sideration the opinions of scholars who have devoted years of 
 study and research to this field. In general, it may fairly be 
 said that she gives insufficient evidence in support of her con- 
 clusions ; sometimes she fails to give any evidence at all.^^ A 
 few quotations will disclose her point of view. On page 178, 
 we read: "As far as direct evidence as to specific African- 
 isms goes, there is nothing to suggest that the Latin of Africa 
 was in any way peculiar". This conclusion is reached in spite 
 of the evidence cited by the author herself, 174-178! Note the 
 
 "But see Brock, ^2. Purser, Ixxiii, makes "Apuleius . . . — at least 
 in the Metamorphoses and the Florida— the most signal representative 
 of the Asianic manner". 
 
 " See below, pages 22-22,. 
 
 '* Studies in Fronto and His Age, 163 ff. I give so much space to 
 Miss Brock's work, not so much because of the importance of her 
 book, as because it is the latest discussion of the subject. The very 
 elaborateness of her treatment might well, in itself, give to her book, 
 in the opinion of the casual observer, a weight which it does not 
 deserve. 
 
 "See e.g. 174, the discussion of the language of Vitruvius. Miss 
 Brock shows here no knowledge whatever of Professor Morris Hickcy 
 Morgan's important studies in the language of \'itruvius. 
 
12 Archaism in Aulus Gellius 
 
 following from 178-179: "It is possible to collect, from the 
 extant writings of African aiUhors, a number of usages, sty- 
 listic and linguistic, which occur first or chiefly in writers of 
 African birth, and some which occur solely in such writers". 
 Statements such as the following on page 182 do not require 
 comment: "In other words, we are to suppose that the Afri- 
 cans who spoke and wrote Latin spoke and wrote it after the 
 manner of Plautus and Cato right on till the second century 
 A.D., and that, therefore, when they began to have a literature^ 
 that literature was naturally archaic. The theory is so im- 
 probable that it would seem almost unnecessary to refute it, 
 were it not soberly put forth by such eminent scholars as 
 Wolfflin and Monceaux, who urge as modern parallels the 
 history of French in Canada or of English in the United 
 States". ^^ Again, on pages 183-184, we read: "One scholar 
 has actually claimed to determine the condition of vulgar 
 Latin at the moment of each provincial conquest by the specific 
 traits of the language spoken in the various countries to-day, 
 . . . The whole theory arose from an erroneous and far- 
 fetched attempt to explain the archaistic movement". Such a 
 theory is, however, still held by competent scholars,^^ and I 
 fail to discover any adequate refutation of it in Miss Brock's 
 book. A final quotation from 184-185 must suffice: "But the 
 greater proportion of archaisms in the conscious archaists, 
 such as Fronto, Apuleius, Aulus Gellius, and Arnobius, found 
 acceptance not because they were still living words on African 
 lips, but because they were sacred by reason of their appear- 
 ance in the old Latin literature, while on the other hand such 
 archaisms as were likewise living vulgarisms were not peculiar 
 to Africa, but formed part of the popular speech wherever 
 Latin was spoken".**' 
 
 Such sweeping statements as the above demand careful veri- 
 fication. No adequate verification, however, appears ; the in- 
 complete lists of words on pages 186-254 present no convinc- 
 ing evidence. Without attempting to prove the contrary, I 
 
 ^ See above, pages 6-9, especially 6-7. 
 
 ^ For recent utterances, see e.g. Grandgent, 2 f. ; Duff, 5 ; for earlier 
 support of this view see Cooper xxviii, and note i. 
 
 ^'Here Miss Brock errs through failure to note the important differ- 
 ences between the first three of the four writers named by her; on 
 these see below, especially 16-17, 21-23. 
 
Archaism in Aldus Gcllius 13 
 
 shall examine the two statements made in the last citation in 
 the light of other statements made by Miss Brock in other 
 parts of her book. 
 
 If I correctly understand what is meant in the clause "But 
 the greater proportion of the archaisms ... in the old Latin 
 literature", it means, in part, that the greater proportion of 
 the archaisms in the authors named by her were not survival 
 archaisms, as I should myself incline to believe in the case of 
 Apuleius, but, if I may use the phrase, revival archaisms 
 culled from the ancients. She has not established the truth of 
 this afifirmation. How does she know that these words were 
 not really current? In view of the scanty literary remains 
 from Africa before Fronto's time, how can we be sure, in the 
 absence of further evidence, that an archaic word found in 
 Apuleius,*^ a writer born and reared in Africa, was taken di- 
 rectly from Plautus or Ennius or other ancient writer, rather 
 than employed in accordance with the usage of the day and 
 country? In other words, how does she know that these 
 archaic words were not in general use in Africa at the time 
 when Apuleius, for example, wrote? How can she be sure 
 that they were used in the period subsequent to Apuleius, be- 
 cause of their "sacred" character and not rather because they 
 W'ere the natural words to use, and had been long current even 
 before his time? Since we have practically no data to argue 
 from, the truth of neither side of the argument can be firmly ^ 
 established. With the scanty data now at hand, it is often im- 
 possible to decide whether an archaic word was borrowed 
 directly from the old writers or had long formed a part of 
 the plebeian vocabulary.*^ If more of the writings of Plautus, 
 
 " Gellius's case is rather difFerent from Apuleius's, in view of his 
 own repeated declarations that he memorized words from the earlier 
 authors for later use; see below, note 106. It will appear below 
 (page 27), also, that it is in vocabulary rather than in syntax that 
 his archaizing tendencies show themselves. See also note 9. 
 
 "Miss Brock, 26, writes: "Indeed, vulgarism and archaism could 
 not but overlap, in view of tlie large proportion of vulgarisms in tiic 
 archaic vocabulary, in the days when the gulf between the written and 
 the spoken language was not so firmly fixed". Reference may be made 
 to note 22 above, to prove, if need be, that there were different kinds 
 even of written language, and that the gulf between one of these kinds 
 and early Latin was not so wide or deep. 
 
14 Archaism in Aulus Gellius 
 
 of Ennius, of Cato, and other old writers were extant, it is 
 probable that we should have to revise our views of what is 
 archaic and what is plebeian. An examination of the Meta- 
 morphoses of Apuleius has disclosed more than fifty words 
 which appear nowhere else in extant literature.*^ Some of 
 them, doubtless, Apuleius himself coined, some probably were 
 words in common use, and others, without doubt, he borrowed 
 from ancient works which are not extant. 
 
 The most surprising statement in the citations given above 
 from Miss Brock's book, however, is in the last clause ("While 
 . . . was spoken": see above, page 12). The process of de- 
 duction does not appear, at least to the present writer, by 
 which she arrives at the broad generalization that all archa- 
 isms which were also living vulgarisms were found in the 
 "popular speech ivherez'cr Latin was spoken". In view of her 
 statement (163), that "it is not the Latin of Africa but the 
 contemporary Latin of other countries which is unknown 
 ground, and it is our ignorance of the latter which obscures 
 our judgment of the former",** her sweeping generalization 
 given above is all the more astonishing. At any rate, she sup- 
 plies no evidence to offset the results of the careful investiga- 
 tions of Cooper, Olcott, and Kiibler, referred to above (page 
 
 9)- 
 
 We are ready now to revert to the remark made above 
 (page 5), that it would be instructive and interesting if we 
 could discover in full the forces at work which produced men 
 of the type of Pronto, Gellius, and Apuleius, etc. It would 
 appear, from pages 5-14, that one force to be reckoned with 
 in the case of at least two of them, is African Latin, or, to 
 put the matter better, the effects of the African environment 
 into which they were born. Apuleius and Pronto, though 
 probably Romans by blood, were of African birth. Apuleius, 
 particularly, gives striking evidence of his southern origin.*^ 
 Pronto aimed at the archaic simplicity of Cato and Ennius,*® 
 
 *'For the new words in the Cupid and Psyche story see Purser, xciii. 
 
 ** Brock, 163 f. 
 
 ** Kretschmann, 4; Cooper, xli. 
 
 *° Kretschmann, 16-17; Purser, Ixxv-lxxvi. 
 
Archaism in Aulus Gellius I5 
 
 but he could not free himself entirely from the forces of 
 heredity and environment.*^ 
 
 A repeated reading of the works of Apuleius, and more 
 especially of the Metamorphoses, has convinced me that the 
 archaic qualities in Apuleius are not, as they seem to be in 
 the case of Gellius, primarily labored imitations of the pre- 
 classical writers,-'^ but, on the contrary, are more often the 
 natural and spontaneous result of heredity and environment*' 
 on an emotional and impressionable temperament/'" Though 
 Gellius's African extraction is only conjectural at best,=' yet iri 
 his case too the African element has to be reckoned with to 
 some extent, indirectly, b-cause he was so deeply influenced 
 by Pronto (see below, pages 16-17). 
 
 The other important factor to be reckoned with in the case 
 of all three, though in varying degrees, is deliberate study of 
 the earher Latin writers and, as a consequence, conscious or 
 unconscious reproduction of the vocabulary, and, far less often, 
 
 *' Mackail, 235 f. 
 
 «Cf. Kretschmann, 17: "Longe aliter Apuleius versatus est, qui 
 quamvis multa et fortasse plura quam primo aspectu videatur antiqua 
 prorsusque obsoleta verba receperit, tamen aliquantum ab orationis 
 prisca quadam specie remotus est. Neque enim ilia antiquitatis vere- 
 cundia eiusque castitatis admiratione commotus, qua Fronto fuit, 
 priscae consuetudinis verba revocavit ; sed ut eorum illecebris orationem 
 docte exornaret, rhetorum praecepta secutus, non aliter atque earn 
 omnibus rhetoricae artis exquisitissimis munditiis distinxit ... Sed 
 si totum dictionis tenorem priscum aut ad vetustatis imitationem adap- 
 tatum esse negamus, eo non infitiamur, totos refertos esse libros 
 veteris consuetudinis vocibus proprietatibusque". Cf. also Piechotta, 
 20-21; Purser, xv, note 2, and, more especially, xciii-xciv. 
 
 *'See Cooper, xviii. xxi-xxvi. with citations; Grandgent, Vulgar 
 Latin 3 Piechotta, 5 ff., has a suggestive discussion of this problem. 
 
 "See Kretschmann, 4-5; 8(fin)-9; Kiibler, Archiv, 8.162.201. In this 
 view the author finds himself in sharp disagreement with the opmions 
 of most scholars who have written on Apuleius (compare, for the 
 current view, e.g. Teuffel § 3671; Brock, 32; Knapp, 135. with the 
 authorities there cited). For another particular in which Apuleuis's 
 style reflects the influence of environment, see below, pages 19-20, on 
 his attitude toward the Greeks. 
 
 "See Teuffel, § 365.1; Cooper, xl, and note 3; Knapp, Stories from 
 Aulus Gellius, 5-6. 
 
i6 Archaism in Aulus Gellius 
 
 of the syntax of those writers (compare above, page 5). 
 Fronto^^ and Gellius^-'' freely avow their devotion to the early 
 Latin writers. Apuleius, not only by certain elements of his 
 own style, but by numerous quotations'^* from earlier Latin 
 works attests his intimate acquaintance with the classical and 
 the preclassical writers. One other special detail may be noted 
 here. Gellius was a student of the old writers on law ;" we 
 may be pretty certain he memorized matters in the course of 
 such reading, as he did, he tells us, in other connections.^* 
 Legal writings" tend toward the use of archaic words and 
 phrases. Their style is inclined to be stereotyped. Gellius's 
 legal training must have given him an additional bias towards 
 the archaic in language. See also below, page 38, on censio. 
 
 These three writers, Fronto, Gellius, and Apuleius, are 
 commonly placed in the same class as the most conspicuous 
 examples of the second century archaisers.^* In this paper, 
 too, heretofore their general likeness to one another has been 
 emphasized. Yet, on closer examination, we find considerable 
 difference between them. Striking as is the agreement, in 
 some respects, in the views of Fronto and Gellius, there are 
 equally striking dififerences. Some of these differences will be 
 noted ; then the striking contrast between Gellius and Apuleius 
 in their attitudes toward the earlier Latin writers will be 
 discussed. 
 
 In the Drisler Studies, 140, we find the following with 
 reference to the mutual relations of Fronto and Gellius: 
 'These passages, few as they are, are sufficient of themselves 
 to make it clear that Fronto exercised a considerable influence 
 
 "See Knapp, Drisler Studies, 135; Kretschmann, 4; Priebe, 6 ff. 
 
 ^ See below, page 17. 
 
 " See below, page 22 ; Vahlen, Ixxx-Ixxxi. 
 
 " Cf. 14.2.1. For Apuleius's legal vocabulary see Helm's Praefatio- 
 to Florida, xii-xiii. 
 
 " Cf. 10.25. 1 ; 11.3.1; 17.2.1; 20.10.4. 
 
 *" Cf. Purser's remark, page Ixxv, that "the jurists, such as Gaius, 
 wrote in a rational way, with a natural leaning towards archaic style, 
 yet not pedantically affecting it". 
 
 "^ Kretschmann, 2; Drisler Studies, 127, 135, 138-140; Cooper, xl; 
 Priebe, 2. 
 
Archaism in Anliis Gcllius IJ 
 
 upon Gellins ; yet they afford no ground whatever for the 
 statement sometimes made that he was a pupil of Fronto. 
 Indeed, our author's words at xix. 8.1 are enough to disprove 
 this assertion". J. Kretschmer, De A. GelHi Fontibus, 103, 
 writes: "Non tarn magister <Fronto> fuit Gellii quam 
 amicus honoratissimus, cuius sermonibus non sine magna utiH- 
 tate adfuisse se fateatur xix.8.1. Scripti Hbri Frontonis Gel- 
 Hus nulHus meminit, neque ego vestigium ullum satis certum 
 deprehendi". 
 
 Professor Knapp states:'"" "Among poets Plautus and En- 
 nius, among orators Cato Censor, stand highest in his (=Gel- 
 lius's) estinlation. Plautus in mentioned or cited by him in 
 at least thirty-five places, and the quotations cover the whole 
 range of the extant plays". ""^ The number of references to 
 Plautus or citations from his plays is even larger than this esti- 
 mate. About forty-five such references and citations are found 
 in the Noctes Atticae. Ennius also is referred to or cited 
 about the same number of times ; Cato's name appears ap- 
 proximately seventy times ; in many instances there are ex- 
 tended quotations. H. Kretschmann"^ does not seem war- 
 ranted in his assertion, at least so far as it refers to our 
 author, that, while Fronto took Ennius as a model. Gellius 
 imitated Plautus, for the w^eight of the evidence is certainly 
 in favor of Cato (see Drisler Studies, 133-134). 
 
 If, however, we were to judge of Gellius's preferences solely 
 by frequency of citation and reference, w^e should have to ex- 
 tend our list of favorites. 
 
 \^ergil would be high up in any such list. The actual num- 
 ber of references and citations is even greater than in the case 
 of Cato.®^ Not one of the twelve books of the Aeneid or of 
 the four books of the Georgics is neglected, and. on the aver- 
 
 " Drisler Studies, 132. 
 
 ** Cf . Stolz. 1.30: "Sicherlich mehr auf Rechnung archaistischer 
 Liebhaberei ist es zu setzen, wenn Gellius vi. 17.4 Plautus also 'homo 
 linguae atque elegantiae in verbis latinae princeps' unci xix.8.6 als 
 'linguae latinae decus' bezeichnet". Cf. also Cooper, x.xxviii. 
 
 " Kretschmann, 16-17. 
 
 "Drisler Studies, 140. See the references in the Index Auctorum, 
 pages 325-326, of Hosius's edition of Gellius. 
 
 ^ 
 
1 8 Archaism in Aldus Gellius 
 
 age, there are four or five citations from each book. There 
 are also three citations (eight lines in all) from the Eclogues.''^ 
 
 Varro is another favorite, and M. Tullius Cicero,*'* by sheer 
 frequency and range of citation, must be placed near the head of 
 the list. The following are the works of Cicero which are re- 
 ferred to or quoted : Brutus, Orator, De Oratore, Orationes 
 in Antonium, Pro Caecina, Pro Caelio, Pro Cluentio, Pro 
 Milone, In L. Pisonem, Pro Cn. Plancio, De Imperio Cn. 
 Pompei, De Provinciis Consularibus, Pro Quinctio, Pro C. 
 Raberio, Pro S. Roscio, Contra Rullum, Pro Sestio, Pro Sulla, 
 In Verrem, Oratio de Accusatore Constituendo, Epistulae®^ ad 
 Atticum, ad L. Plancum, ad Ser. Sulpicium, De Amicitia, De 
 Divinatione, De Fato, De Finibus, De Officiis, De Republica, 
 De Gloria, De lure Civili, Tusculanae Disputationes. 
 
 In view of the incomplete and fragmentary character of the 
 extant works of Fronto, it would not be safe to attempt to 
 draw too definite conclusions concerning his favorite Latin 
 authors. The Index Scriptorum in Naber's edition (268-270) 
 includes references to citations from these Latin authors, be- 
 side others: Accius (four citations), Caecilius (two), C. 
 Caesar (three), Cato (twenty- four), Cicero (twenty-two), 
 Ennius (fifteen), Lucretius (five), Naevius (three), Plautus 
 (seven), Sallust (fourteen). Doubtless, if more of Fronto's 
 literary works had survived, they would only emphasize the 
 facts just stated. Cato and Ennius are in high favor. Plautus 
 is in fifth place. We do not find Vergil and Varro in Naber's 
 list. 
 
 "On Vergil's influence upon the literature of the centuries since his 
 time see e.g. Knapp, Originality of Latin Literature, The Classical 
 Journal 3.252; Sellar, Virgil, 60 ff. 
 
 " Drisler Studies, 130. It is to be noted that Cicero was rated by 
 Gellius, as by others before him, among the antiqui and the veteres; 
 see below, page 30; Drisler Studies, 129-132. Hence Gellius's 
 frequent citation of Cicero is after all but one more proof of his 
 archaizing. Varro's antiquarian tendencies in scholarship and re- 
 search need no illustration beyond a reference to Teuffel, 164. 2. 
 In language, too, Varro was an archaist ; see e.g. Schmalz, 665. For 
 the reading of Varro in Gellius's time see the Prologomena to A. 
 Riese's edition of Varro, Saturae Menippeae, pages 51-52. 
 
 *'For Fronto's opinion of Cicero's letters see below, page 19. 
 
Archaism in .hilus Gcllins 19 
 
 In this connection it is worth while to note the difference in 
 attitude of Fronto and GelHus toward Cicero's letters. An- 
 toninus writes (Epistulae 2.4: Naber, page 107) to Fronto 
 as follows : Ciceronis epistulas, si forte electas, totas vel dimi- 
 diatas habes, imperitas.vel mone quas potissimum legendas mihi 
 censeas ad facultatem sermonis fovendam. Fronto replies 
 (2.5) : Memini me excerpisse ex Ciceronis epistulis ea dum- 
 taxat quibus inesset aliqua de eloquentia vel philosophia vel de 
 republica disputatio : praeterea si quid eleganti aut verbo 
 notabili dictum videretur, excerpsi. Quae in usu meo ad 
 manum erant. misi tibi. . . Omnes autem Ciceronis epistulas 
 legendas censeo mea sententia vel magis quam omnes eius 
 orationes. Epistulis Ciceronis nihil est perfectius. 
 
 Gellius mentions Cicero's letters only three times (1.22. 19; 
 4.9.6; 12. 13. 21). In the first instance, merely to illustrate a 
 certain use of supcrcssc, he quotes from Ad Fam. 10.35.5 • 
 Nam neque deesse rei publicae volo neque superesse. This is 
 quite in accord with the Gellian manner. With a like purpose 
 in mind, that is, to define the phrase dies rcligiosus, he makes 
 his second quotation, from Ad Att. 9.5.2. It is to comment on 
 the phrase intra modnm that the third quotation (Ad Fam. 
 4.4.4) is given. In none of the three passages does he utter 
 one word of formal or special commendation of Cicero's let- 
 ters, prone as he was to eulogize in set terms those from whom 
 he cites. From all this, however, we may hardly be justified 
 in the inference that Gellius did not regard the letters highly, 
 but it is at least safe to conclude that he did not find in them 
 the material suited to his needs. 
 
 In the Drisler Studies, 140, attention is called to Fronto's 
 attitude toward the Greeks as contrasted with that of Gel- 
 lius.*" Apuleius shared with Gellius his admiration for the 
 
 "See Gellius, 1.8.6; 10.22.3; 11.16.1,9; 12. 1.24; 14.1.32; 15. 11. 3; 
 17.20.7,8; 18.13. 5. Two of these passages may be cited in full: 10.22.3 
 Verba . . . Platonis . . . scrips!, quoniam vertere ea consilium non 
 fuit, cum ad proprietates eorum nequaquam possit Latina oratio aspi- 
 rare ac multo minus etiam mea: 12. 1.24 Haec Favorinum dicentcm 
 audivi Graeca oratione. Cuius sententias communis utilitatis gratia, 
 quantum meminisse potui. rettuli, amoenitates vero ct copias ubertates- 
 que verl)orum Latina omnis facundia vix quadam indipisci potuerit, 
 
20 Archaism in Aulus Gcllius 
 
 Greeks. A glance at the Index Auctorum in Helm's edition 
 will be sufficient to prove this interest. Almost every page of 
 the Florida reveals the name of some Greek poet or philo.-- 
 opher. Apparently one of the accusations against Apuleius 
 v^^as the charge that he was equally a master both of Latin 
 and of Greek (Apol. 5.5).*'' Fronto, on the other hand, de- 
 spised, or affected to despise, the Greeks : see e.g. Drisler 
 Studies, 140; Brock, 38, 41.''^ 
 
 Although Gellius's chief interests lie mainly with the early 
 classical and the preclassical writers, he was by no means ignor- 
 ant of the literature of the first century. However, when one 
 remembers the great number of Latin authors mentioned by 
 him (I have noted well over seventy), his silence in regard to 
 Propertius and TibuUus, Livy and Tacitus, Quintilian and 
 Juvenal, is surprising."^ The nearer he approaches his own 
 time, the less he has to say about literary men and their 
 works.'" Still, references to the writers named below attest 
 some knowledge on his part of the literature from the opening 
 of the Christian era to his own day: Augustus Caesar 
 (10.11.5; 15.7.3) '■> L. Annaeus Seneca (12.2.3 ff.) ;'^^ C. Ateius 
 Capito (1.12.8) ; Valerius Maximus (12.7.8) ; Valerius Probus 
 (1.15.18) ; Q. Asconius Pedianus (15.28.4); Suetonius Tran- 
 quillus (97-3; 1544) • 
 
 Of Gellius's contemporaries many are named; some of these, 
 with others unnamed, share Gellius's love for the early Latin 
 writers. Fronto'^- especially is looked upon as an authority in 
 
 mea tenuitas nequaquam. For references to Gellius's visit to Greece 
 see 1.2.1; 2.2.1,2; 7.13.1,2; 7.16.1; 8.10; 9.4.1; lo.i.i; 12.5. i ; 15.2.3; 
 15.20.5; 16.6.1 ; 1 7.8. 1 ; 18.2. 1 ; 18.9.5; 18.10.3; 18.13. i : 19.1.1,4; 19.6.2; 
 18.8.1 ; 19.12. 1. 
 
 " Yet, as Kretschmann, 67 ff., and Purser, xcv, note, Apuleius uses 
 few Greek words. For the wide use of Greek in Africa in Apuleius's 
 time see e.g. Purser, xvi, note. 
 
 " Purser, Ixxi, disregards this attitude of Fronto. 
 
 ^ Drisler Studies, 134. 
 
 " Teuffel, § 365.5. 
 
 " But Gellius names Seneca chiefly to criticize him severely. Fronto, 
 too, shows a decided antipathy to Seneca; see Teuffel, § 298.1. 
 
 "See above, 5; Drisler Studies, 139-140. 
 
Archaism in Axilns Gcllius 2i 
 
 matters linguistic (2.26; 13.29; 19.8.10-13). Apuleius is not 
 mentioned, although it is possible that his student days at 
 Athens" fell at about the time when Gellius was in that 
 city/* The standards and aims of Gellius and Apuleius were 
 so diflferent that the former's silence with respect to his more 
 gifted contemporary should cause no surprise (cf. page 20, 
 above). 
 
 It is interesting now to compare Apuleius and Gellius in 
 respect to their citations from Latin authors. Gellius is par- 
 ticularly interested in the subject-matter of his citations. The 
 ipse dixit of an Ennius, a Cato, or a Varro settles beyond fur- 
 ther argument any point relating to customs or language (see, 
 for example, 16.14 or 18.9). Gellius seems to have mainly ^ 
 two objects in view: first, to display his erudition, and, sec- 
 ondly, to instruct." His quotations have a practical purpose, 
 not an aesthetic." Not the beauty of a passage, the imagin- 
 ative, the literary qualities, appeal to our author. He quotes a 
 passage because it has an unusual word or form in it, or an 
 allusion to some old law, to some quite- or half -forgotten cus- ^ ^ 
 tom, or for some other similar reason which appeals to his 
 prosaic grammarian's taste (see, for example. 1.7. 11; 1.11.16; 
 1.16.1-5; 1. 18.2; 1.21; 1. 25. 17; 2.14; 19.8). 
 
 Apuleius has quite a different purpose in view. He is the 
 popular lecturer, the elegant raconteur, the eloquent and gifted 
 orator, the learned expounder of a mystic philosophy. He is 
 interested in making his point, in entertaining his readers, in 
 dazzling his auditors. '^^ His citations are mere literary orna- 
 
 "' Apuleius refers to his sojourn at Athens in Met. (Helm's edition: 
 references are to page and line) 4.2; 22.12; Ap. 80.16: Flor. 35.15; 
 39.6. See Purser, xii-xiii. 
 
 "For Gellius's references to his stay at Athens see above, note 66. 
 Purser, xiii, holds that Apuleius remained at Athens "probably 
 . . . till 150 A.D. ; possibly he remained later". Knapp, Stories from 
 Aulus Gellius, 7, showed that Gellius's sojourn in Athens antedated 
 166 A.D., though by how many years he did not venture to say. 
 
 " Praefatio, 2, 10, 12-13. 
 
 "See, however, 19.11: 1.24. 
 
 "If Purser, Ixix-lxxxiv, is right in regarding Apuleius as an 
 "Asianic" in style, his remarks arc in point here, especially Ixxvii. 
 
22 Archaism in Auliis GcUius 
 
 ments, employed for their rhetorical effects. Apuleius is no 
 grammarian, absorbed in the consideration of the minutiae of 
 language. He is a literary artist, profoundly affected, to be 
 sure, by his predecessors who wrote two and three centuries 
 before his day, but still mingling, with certain archaic ele- 
 ments, those qualities of vocabulary and style which were, 
 doubtless, peculiar to his own native Africa. '^^ The same 
 richness of vocabulary, the same ornate and florid style are 
 found in Tertullian,'^'' and, to a less degree, in Cyprian. *° A 
 few examples will illustrate what was stated above (page 21) 
 with reference to his citations, that they are literary orna- 
 ments: see Apol.^^ 14.16-15.1 eleganter Afranius hoc scriptum 
 relinquat: amabit sapiens, cupient ceteri ; 6.23-7.1 sane quidem, 
 si uerum est quod Statium Caecilium in suis poematibus scrip- 
 sisse dicunt, innocentiam eloquentiam esse, ego uero profiteor 
 ista ratione ac praefero me nemini omnium de eloquentia con- 
 cessurum; Flor. 3. 15-17 prorsus igitur ante Hyagni(n) nihil 
 aliud plerique callebant quam Vergilianus upilio seu busequa, 
 'stridenti miserum stipula disperdere carmen' ; 42.7-9 equum 
 deligunt diutinae fortitudinis, uiuacis pernicitatis, id est et 
 ferre ualidum et ire rapidum, 'qui campos collesque gradu 
 perlabitur uno', ut ait Lucihus. These illustrations could be 
 multiplied almost indefinitely. 
 
 The actual range of authors cited by Apuleius and by 
 Gellius, while by no means identical, is the same to a re- 
 markable degree. Both cite Accius, Aedituus, Afranius, Cae- 
 cilius, Calvus, Cato, Catullus, Caesar, Cicero, Ennius, Horten- 
 sius, Lucilius, Lucretius, Plautus, Sallust, Varro, Vergil. The 
 number of Latin authors, however, cited by Gellius (I have 
 noted approximately seventy) is much greater than that cited 
 by Apuleius (approximately twenty). It is not so much, then, 
 in the range of authors cited as in the motives that inspire 
 citations that Gellius and Apuleius differ. 
 
 " Stolz, 49. But see E. Norden, 588-598 (his views are stated and 
 discussed above, note 27) ; Purser, Ixxxiv ; Brock, 163 ff. Com- 
 pare Teuffel § 366.6. See above, pages 6 ff. 
 
 " Mackail, 252. 
 
 '" Mackail, 255. 
 
 "Helm's edition (page and line are cited). 
 
Archaism in Aldus Gcllius 2^ 
 
 They differ again in the frequency of citations. Apuleius 
 cited Ennius four times, Plautus four times, and Cato only 
 three. It is interesting to note that both Gellius and Apuleius 
 have one reference to Cato in common.^- Varro is cited by 
 Apuleius twice, X'ergil ten times. Apuleius rarely praises these 
 authors whom he cites.^^ This is in sharp contrast to Gel- 
 lius's frequent laudations of the old writers.*** 
 
 It was argued above (pages 9-1 1) that the archaic element 
 in Apuleius is in large part the inevitable outcome of his 
 African birth. It was remarked, however, that this view has 
 not been accepted by all scholars. Some attention will now 
 be given to the more commonly accepted theory. 
 
 We may do this most conveniently by considering briefly 
 some points discussed in Leky's De Syntaxi Apuleiana (1908). 
 Leky regards it as an established truth that Apuleius is, pri- 
 marily, an intentional archaist^" and so throughout his disser- 
 tation charges all the archaic elements of Apuleius's syntax to 
 direct and conscious imitation. Desertine^** takes the same 
 ground with reference to both vocabulary and syntax.®^ 
 Kretschmann (pages 34 ff. and 87 ff.) holds the same view. 
 Leky's work consists in comparing certain syntactical usages 
 found in Apuleius with similar uses in the preclassical period, 
 particularly in Plautus. His examples do show striking re- 
 semblances. It is suggestive to note, however, that while 
 Apuleius and Gellius often exhibit similar archaic qualities 
 of syntax, as will be shown farther on. these qualities are more 
 conspicuous in Apuleius ; still, in him, they seem quite natural 
 and unaffected.^^ If we may consider Met. i.i to refer to 
 
 ** Gellius, 2.2.28; Apul. De Mundo, page 150.7. 
 
 ^'But see De Deo Soc. (Thomas's edition), page 18.12, and Apol., 
 page 14.17. 
 
 " See above, note 41 ; below, note 106. 
 
 "Leky, 5-6. 
 
 "Cf. Desertine, i; Duff, 30: and the authorities cited in Drisler 
 Studies, 135. 
 
 •' Pages 81 ff. 
 
 "Desertine, 81. Reference may be made again to the argument on 
 pages lo-ii above, and to note 41. For the comparatively small 
 amount of archaism in syntax in Gellius see below, page 2j. 
 
24 Archaism in Aldus Gellius 
 
 Apiileius himself, we shall conclude that he even exerted him- 
 self to avoid the very things in which Gellius delighted. 
 
 Leky,*^ commenting upon the omission of the finite verb in 
 Apuleius, after giving examples of such omission in Plautus, 
 remarks : "Huiusmodi igitur apud Apuleium dicendi rationem, 
 quamvis etiam aliunde orta esse possit, tamen pro summo 
 Apulei veterum studio et ostentatione non sine bono iure ex 
 priscorum imitatione repetemus". He cites Haupt, Op. 3.377, 
 for examples from Petronius and Phaedrus, as well as Mad- 
 vig's discussion of this usage in Cic. De Fin. 1.9. On page 
 8 again, Leky, with reference to the same usage, writes : 
 "Haec exampla . . . quamvis ne ab aliis quidem scriptoribus 
 prorsus sint aliena. . ." This admission materially weakens 
 his claim that such usages were mere imitations of the an- 
 cients. They are evidently plebeianisms. 
 
 An adverb modifying a substantive is found in Apuleius. 
 The same usage is found in Plautus. Leky notes^** that 
 Cicero rarely employs it. Livy, Tacitus and the Augustan 
 poets employ it more frequently. Yet Leky concludes (page 
 11): "ex Apuleianis tamen, quae mihi certe videntur ad 
 comicorum sermonem addecere, sunt". On page 36, we find 
 another illustration of Leky's method, or rather of his in- 
 ability to see the implication of his own citations: "Indica- 
 tivi"^ usus apud Apuleium multo liberior est quam apud op- 
 timae latinitatis scriptores. Atque banc dicendi libertatem ab 
 Apuleio plerumque ex priscorum sermone sumptam esse ex 
 €xemplis sequentibus elucebit". He mentions the indicative in 
 dubitative questions as an example of this imitation of the 
 ancients, and then directly says : "Quae dicendi ratio . . . 
 vulgi sermoni propria est". He notes (page 36) its use in 
 letters of Cicero ad Atticum, in Catullus, in Vergil. On page 
 37, in regard to certain uses of the subjunctive, he says: 
 "Exemplo certe comicorum Apuleius facit, ut alteram con- 
 iunctivi personam numeri singularis ponat pro imperativo, 
 quem priscorum usum poetae, ut Catullus, iam receperant, 
 
 '' Leky, 6. 
 
 "'Leky, II. 
 
 " See Brock, 195. 
 
V 
 
 
 Archaism in Aiilus Gcllius 25 
 
 quemque interdum invenimus in Cic. epp". The above quo- 
 tations are intended to show that Leky takes it for granted 
 that, when Apuleius's language differs from the classic norm 
 so as to show an archaic quality, it is because he deliberately 
 imitates the ancients."- He gives very little weight even to his 
 own citations of works in which the plebeian elements are 
 generally acknowledged, and which, being of a later date than 
 Plautus, tend to show that such usages in the Latin language 
 had maintained themselves side by side with the literar>' 
 speech, during the classic period, even at Rome. We should 
 expect to find that, in the Latin language as it was spoken and 
 written in Africa,"^ the archaic-plebeian elements would be 
 striking. A more illuminating study of the language and 
 style of Apuleius might be made by comparing him, not only 
 with Plautus and other early writers, but with writers who 
 were of, or near, his own time, especially with others of the 
 African school, such as Tertullian and Cyprian. If these 
 same archaistic qualities should be found in these writers, who 
 could hardly be charged with intentional archaising,''* it would 
 be a fair inference that Apuleius, also, did not borrow so much 
 directly from Plautus as he is generally believed to have bor- 
 rowed. 
 
 Although he appears to give little importance to it, Leky, by 
 his own investigations, proves that many of the very qualities 
 of Apuleius's syntax which he has ascribed to conscious imi- 
 tation of Plautus are found in the later writers, among whom 
 are many who are not generally regarded as archaisers but 
 whose works are recognized as being tinged with a distinctly 
 plebeian color ;^''' we may name here Auctores Belli Africani 
 
 "But cf. Flor., page 10.20 ff. ; Piechotta, 3; Kretschmann, 4. 
 
 "Cooper xviii, xxvii ; Piechotta, i flf. ; above, pages 5- 14. 
 
 "Compare Ott, in Neue Jahrbiicher, 109. 762: "Fachschriftsteller 
 wie die arzt Caelius Aurelianus und der theologe Tertulliainis, denen 
 es doch vvolil nicht um rhetorische effecthashcerei zu thun ist, zeigen 
 in beiden stuecken die ganz gleichen erscheinungen. Und wie oft 
 gemahnen die alien uebersetzungen der Bibel und des Irenaeus an 
 Plautus und die archaische literatur ueberhaupt". But see Brock, 
 182-184. 
 
 " Lel<y, 15-29. 
 
26 Archaism in Aulus GclHus 
 
 et Hispaniensis, Vitruvius, Phaedrus, Petronius, and Tertul- 
 lian. On page 21, he attributes to imitation of Plautus Apu- 
 leius's use of the infinitive depending upon a verb of motion to 
 express purpose. Even though parallels of the so-called ar- 
 chaisms of Apuleius are found in Cicero's letters, in Quintil- 
 ian,^" and may be produced from the poets and the writers of 
 plebeian Latin, Leky still maintains his thesis that Apuleius 
 borrowed directly from the veteres, and particularly from 
 Plautus. From page 31 I take the following, which epito- 
 mizes Leky's mental attitude: "facile commovemur, ut a pris- 
 cis sumptum esse usum suspicemur". 
 
 To the present writer, Leky's dissertation proves that the 
 syntax of Apuleius has certain qualities which are found not 
 only in Plautus, but also in the sermo plebeius of later periods. 
 Cooper^^ and Piechotta®^ prove conclusively that, so far as 
 word formation and vocabulary are concerned, the Metamor- 
 phoses of Apuleius, particularly, furnishes a notable example of 
 plebeian Latinity (see above, pages lo-ii). 
 
 That Apuleius abounds in archaic and rare words no one 
 who has examined the Metamorphoses will be inclined to 
 deny. But the same may be said of Tertullian and of Cyp- 
 rian.^^ The African inscriptions also show a similar vocabu- 
 lary.^*'" Leky has failed to prove that the archaic tinge to 
 Apuleius's syntax is due to intentional imitation. He has, on 
 the contrary, shown that, in a very large number of instances 
 wherein Apuleius's syntax differs from the classic norm, it is 
 plebeian. A more detailed comparison with the syntax of 
 
 "Leky, 30. 
 
 " xl ff., and lists. 
 
 " 25-52. 
 
 ^'Cf. examples in Piechotta, 28 f., and Cooper xxvii. The latter 
 .says: "especially notable is the strongly archaic element in the sermo 
 Africus, which, as seen in Fronto, Apuleius, Tertullian, etc., presents so 
 many striking analogies with the language of Plautus. This phenome- 
 non is easily accounted for, when we remember that the first germs 
 of Latin were carried to Africa by the Roman soldiers and colonists 
 who flocked there after the fall of Carthage, in 146 B. C". See also- 
 Cooper's footnotes ; above, page 9. 
 
 "" Cf. Kiibler, Archiv, 8. 201 f. 
 
Archaism in Aiiliis Gcllins 27 
 
 TertulHan and Cyprian would probably have shown that /\pul- 
 eius's syntax did not especially differ from the usages of other 
 African writers of about his own time. 
 
 The archaic element is not so obvious and striking in Gel- 
 lius's syntax as in his vocabulary. ^''^ It is not difficult to find 
 an explanation for this fact. To intersperse one's writings 
 with obsolete and archaic words may indicate profound ad- 
 miration for the writers of the older days and a wide ac- 
 quaintance with their works (an acquaintance which may, 
 however, be after all somewhat superficial), yet does not 
 necessarily involve that instinctive feeling which enters into 
 the very spirit of the ancients, which a successful imitation of 
 their syntax would imply. In general, then, Gellius uses his 
 moods, tenses, cases, and arranges the order of his phrases and 
 sentences, not after the fashion of Cato, but according to the 
 canons of his own time. However, here and there we dis- 
 cover a case construction, a use of a preposition, an em- 
 ployment of a mood, or the turn of a phrase, which smacks of 
 a period three centuries before Gellius's day. 
 
 As in the case of his vocabulary, so even more in the matter 
 of syntax, one cannot always be sure whether these peculiar 
 usages are the results of conscious imitation, or are uncon- 
 scious reminiscences of the old authors whose works Gellius 
 conned so assiduously, and from which he habitually memor- 
 ized passages, or are merely plebeianisms which had gained 
 currency among the cultivated classes of his day. 
 
 Gellius comments freely upon the uses of words, some- 
 times defending the archaic as against the popular. In more 
 than one instance he employs words which elsewhere he has 
 declared to be obsolete.^"- Purist he claims to be, but a purist 
 in language, from his point of view, looked for authority not 
 to the usages of his own day but to an Ennius or a Cato. He 
 
 "'From this point on, the argument on pages 5-1 1 above, that in 
 Apuleius we have a large element of 'survivar archaisms, whereas in 
 Gellius we have rather, in the main, conscious 'revival' archaisms (aside 
 from those cases where the plebian and the archaic coincide) should 
 be kept steadily in mind. See again above, note 41. 
 
 ""Drisler Studies. 146. 
 
28 Archaism in Aulus Gellius 
 
 seldom comments upon syntactical matters. ^''^ He may slip in- 
 to some sentence an apparently obsolete construction, but he. 
 usually does so unostentatiously, perhaps unconsciously, with 
 no comment, and with no citation of ancient authorities."* 
 
 In his paper entitled Archaism in Aulus Gellius,"' Pro- 
 fessor Knapp undertook to determine the exact extent of Gel- 
 lius's indebtedness to anteclassical writers. In the first part 
 of the paper (126-141), by means of citations from the Noctes 
 Atticae, Gellius's devotion to the vetcres, antiquiores, maiores 
 was set forth. From the testimony of other authors, and from 
 evidence furnished by Gellius himself, the meaning to be given 
 to these terms is made evident. Finally, the author traces the 
 rise and growth of antiquarianism in Latin Literature to its 
 culmination in the second century of our era. 
 
 In the second part (141-171), some archaisms of form were 
 noted, and then were given alpabetical lists of nouns, adjectives, 
 verbs, adverbs, prepositions and conjunctions, which Gellius 
 was believed to have derived from older Latin writers. 
 
 We shall now proceed to give additional references and au- 
 thorities in connection with the words already treated in the 
 Drisler Studies. A few archaisms of form and vocabulary 
 which were not noted there will be added. 
 
 "'See, however, 1.7. 13; 17.2. 11. 
 
 ^*'* Contrast his utterances about archaic words : see below, page 29. 
 
 "° Printed in Classical Studies in Honour of Henry Drisler, 126-171.- 
 
PART I 
 
 Archaisms of Form and Vocabulary 
 
 A. Forms. 
 
 To the passages cited on page 128 to show GelHus's love for 
 the older writers may be added 17.2. i : Cum librum veteris 
 scriptoris legebamus, conabamur postea memoriae vegetandae 
 gratia indipisci animo ac recensere, quae in eo libro scripta 
 assent in utrasque existimationes laudis aut culpae adnotamen- 
 tis digna, eratque hoc sane quam utile exercitium ad conci- 
 liandas nobis, ubi venisset usus, verborum sententiarumque 
 elegantium recordationes. In addition to the statement that he 
 habitually committed to memory words and phrases of older 
 writers/*'*' we have the avowal that in the vctcres scriptores 
 perfection was^°^ to be found. It is worth while to note that, 
 of the expressions which, according to 17.2.2, he had memori- 
 zed from Claudius Quadrigarius, there occur elsewhere in the 
 Noctes Atticae, employed by Gellius without reference to their 
 source, cnmpnmis (1.13.7; i-i5-8; ii-3-i I I3-I7-2; 13-21.25 ; 
 18.4.8; I9-5-3);''" ne . . . quoqiic (1.2.5; ii-54; 20.1. 15) ; in 
 medium relinquo (7.14.9).^**^ According to GelUus himself, 
 cumprimis and ne . . . quoquc were obsolete or nearly so 
 (17.2.14,18).^^° It is interesting to note that the regular form, 
 ne . . . quidem, is used by Gellius at least twenty-three times 
 and that three of these occurrences (i 1.5.8; 20.1. 14; 20.1.33) 
 are removed by only a few parapraphs from the three instances 
 of the archaic nc . . . quoqiie. 
 
 In referring to the older writers,'" Gellius usually char- 
 acterizes them as antiqui, antiquiorcs, or maiorcs, or states 
 that a word or construction was used antiquitus. In like man- 
 ner, in 247 out of 278 passages in the two versions of Servius's 
 
 "•Cf. 10.25. 1 ; 1 1. 3.1; 20.10.4. 
 
 "'Cf. 17.2.6; 17.2.10. 
 
 "* See below, page 36. In 13. 17.2 he is talking of the veteres. 
 
 '"* See below, page 55. 
 
 "• Drislcr Studies, 170. 
 
 "'Cf. Drisler Studies, 128-129. 
 
30 Archaism in Aulus Gelliiis 
 
 Commentary on Vergil in which reference is made to the ar- 
 chaic character of the words used by Vergil, the author employs 
 the words antiqid, vetcrcs, maiorcs, or some form of these 
 words/^- 
 
 On pages 130-132 of the Drisler Studies, the meaning of 
 the terms veteres, antiqui, and the like is discussed. Compare 
 with the citation from Ouintilian 9.3.1, which clearly refers to 
 Cicero and his predecessors, the following from Gellius him- 
 self : 13. 17.2 Sic igitur eo verbo veteres esse usos et cumprimis 
 M. Varronem"^ Marcumque Tullium omnes ferme libri de- 
 clarant; 9.12.4 Sed et veteres plerique ita dixerunt, et M. 
 Tullius. 
 
 To the list of writers (132) by whom the plural of arena 
 is employed add Vergil, Horace and Columella."* To the 
 notes (141-146)"^ on archaisms of form may be added aecum: 
 16.4. in lemm. For prorsiis and prorsum are found the archaic 
 syncopated forms prosus and prosum.^'^^ For the former see 
 2.8.7; 2.22.25; 4.13.4; 10.3. 10; 14.6.5; 20.5.8; 20.5.10. See 
 also below, pages 45-46. Both forms are used by Fronto in his 
 correspondence.^^^ 
 
 Temperi/^^ used in 2.29.1 1, Quin potius imius et cognates 
 adfinesque nostros oramus, ut assint eras temperi ad meten- 
 dum, is probably an archaism ;"^ this chapter bristles with 
 archaisms. ^^° Hosius reads temere in praef. 18, where Hertz 
 reads tempere}-'^ 
 
 Aliqiii is used as a substantive: cf. 11. 13.5 inspicite penitus 
 quid efificiant verba haec, dicatque mihi, quaeso, aliqui vestrum 
 
 ^See Steele, Archaisms in Vergil, A.J.P., 15.166. 
 
 "^ Compare above, note 64. 
 
 "* See Lewis and Short, s.v. 
 
 "^ Drisler Studies. 
 
 "° See Georges, s.v. ; Neue-Wagener, 2.747. 
 
 "' See Neue-Wagener, 2.747. 
 
 "" Cf. Knapp's paper on Vahlen's Ennius, A.J. P., 32.29, note. 
 
 "* See Georges, s.v. ; Neue-Wagener, 2.649 ; and Lorenz on Plautus 
 Pseud. 375. 
 
 ^^ Knapp, A.J. P., 32.29, note i. 
 
 ^ See Hertz, Vindiciae Gellianiae Alterae, 26. Hertz took tempere 
 as a by-form of temperi: see Neue, Lexicon der lateinischen Wort- 
 
Archaism in Aldus Gcllius 3^ 
 
 an sit ulla huiusce sententiae gravitas aut gratia. According to 
 the references given in Neue-Wagener, 2.476, the usage is 
 rare, beginning with Caesar and Cicero. For the first century 
 and a half of the empire but four citations are given, two from 
 Seneca's Epistles, one from Livy, 4.35.9, one from Quint., 
 Decl. 294. It would seem that the use never gained 
 ground, and was in part obsolete by Gellius's time.^-- Gellius 
 derived it then, from his reading of the vetcrcs}-^ 
 
 It may be noted here that Hertz in his editio altera minor 
 (1886) reads quo in 15. 10.2 instead of the qui of his critical 
 edition (1885). Hosius reads quo. In addition to the com- 
 ments on qui in the Drisler Studies (145), there should be 
 noted the use of both qui and quo side by side in 1.13.11. The 
 frequency with which quo occurs elsewhere in Gellius makes 
 these isolated examples of qui all the more striking. 
 
 It should be noted that there are some traces in the MSS of 
 Gellius of an archaic form of the dative of the third declension 
 in -e}^^ The forms are parte, 2.12.1, and corpore, 3.1. 13. 
 Hosius and Hertz both read parti and corpori, but Hertz holds 
 that the -e forms are not impossible.^-"' 
 
 Viderier occurs also in 1 5.2.1 Ex insula Creta quispiam . . . 
 sese philosophum dicebat et viderier gestibat. Note the two 
 archaisms together here, viderier, and gestibat. For the latter 
 form, add to the references on page 146 Munro on Lucr. 5.934, 
 
 formen, s.v. tempore; Biicheler, Rhein. Mus., 15-444; Neue-Wagener, 
 
 2.649. 
 
 '^ See the note in Hildebrand's edition of Apuleius on Met. 6.29 
 (Vol. 1.504); Neue-Wagener, 2.476; Thesaurus, s.v. 
 
 ^I use the term here in Gellius's understanding of it (see above, 
 pages 29-30). 
 
 "* Lindsay, The Latin Language, vi. § 28, page 387. 
 
 "'See Vind. Cell. Alt., 8; Gorges, 18; Bucheler. Grundriss, §§ 276- 
 278. J. Gronovius ad loc. declared that a collation of a certain MS 
 which he possessed gave iure dkundo in 13. 12.9; this reading, he ad- 
 ded, "est vere Romanum". But there Hertz had iuri dicundo (so too 
 Hosius, with no comment whatever in his apparatus criticus). In his 
 Vind. Gell. Alt., 8, Hertz declined to put this passage beside 2.12.1, 
 3. 1. 13, as evidence for a dative in -e. because Gronovius's reading "auf 
 den cod. reginae in Vat. 597 zuruckzugehen scheint, bei dem ein 
 falsches in iuri dicundo nachtriiglich in in iure dicundo verbessert ist". 
 
32 Archaism in Aulus Gellius 
 
 and Steele on lenibat, A. J. P., 15.188. Fervif'^ appears also 
 in 17.8.8. 
 
 To the forms of queo cited on page 146 add the following: 
 queunt 14.1.30, 16.8.14, 16.13.9; quire 11.9.1; queat 2.6.9, 
 12.12. 1 ; queant 14.1.26 (bis) in affirmative clause; quiret 
 14.1.12; quiverit (subj.) 14.1.17 in affirmative clause; quivi 
 14.2.25. Note also nequivit 6.3.3; nequirct 2.23.22. 
 
 An archaic form not noted in the Drisler Studies is par- 
 sisset,^-^ 5. 14.15. In 15.31. 5, however, Gellius writes pepercit. 
 
 Worthy of a place here, perhaps, is defio, a by-form of de- 
 ficio, which appears twice: 1.14.1 quod viderent multa ad 
 splendorem domus atque victus defied ; 20.8.5 deficiente contra 
 (luna) defiunt. In the latter passage note deficio and defio 
 side by side. Defio is found in Plant., Ter., Enn., Ace , Lucr., 
 Verg., and later poets, but in prose apparently only in Livy,^^* 
 Vitruv., Gell., and later. Gellius shows no knowledge, it may 
 be noted, of either Livy or Vitruvius (see above, page 20) ;. 
 it is likely, therefore, that he derived the forms under discus- 
 sion from his beloved veteres. 
 
 B. Vocabulary. 
 
 We shall proceed now to add such comments and references 
 as seem profitable to the words Gellius is believed to have bor- 
 rowed from the older writers. This list of words begins in 
 the Drisler Studies with page 147. Since the words are ar- 
 ranged alphabetically, it will be unnecessary to cite the page 
 on which the word is treated in Drisler Studies. 
 
 acritudo: for the archaic character of the ending -tiido see 
 Cooper, 44, and Schmalz, 604 ( § 2a ) : "Die Neigung zur 
 Abstraktion in den Zeiten des Verfalls zeigt sich ... in 
 Wiederaufnahme der . . . von den Klassikern vernachlassigten, 
 . . . aber bereits von den archaisierenden Schriftstellern der cic. 
 
 "'See Georges, s.v. ferveo; cf. Vergil Aen. 4.409. 
 
 ""See Georges, s.v. parco; Steele, A.J. P., 15.188. 
 
 "'See Weissenborn-Mueller on 9.11.6; Weissenborn on confieret, 
 5-50.7- 
 
Archaism in Aulus Gcllius 33 
 
 und der folgenden Zeit gerne aufgegriffenen Bildung mit tudo, 
 z. B. claritudo". 
 
 canalicxda, 1 7.1 1.2: the statement that this word is found 
 only in LuciHus and Gellius requires correction. It occurs also 
 in Varro.^-® 
 
 finis, in 4.1.6 signifies 'meaning' rather than 'termination' 
 (correct, then, Drisler Studies, 150). Postgate, Select Elegies 
 of Propertius, xci, remarks: "His genders are sometimes 
 archaic: e. g. ptiluis and fi)iis are common..." Finis is 
 masculine in 1.3.8 and 3.16.1, but feminine again in 3.16.20. 
 
 lac: lade is read as the nominative of this word in 19.8. 13 
 cur mel et vinum atque id genus cetera numerum multitudinis 
 capiunt, lacte non capiat? With what is said under this word 
 in Drisler Studies concerning the identity often of archaisms 
 and vulgarisms compare Woliflin, Philologus, 34. 149.'^° 
 
 partio: add 12. 1.20, and Macrob. 5. 11. 15 (borrowed, as is 
 so much in Macrobius, from Gellius). 
 
 specus is feminine in 5. 14.18; the word does not occur at 
 all in 5.14.8 (correct Drisler Studies, 153). Add 5.14.24 in 
 cadcm specu. The word occurs in still another place, but 
 there the gender is not clear; see 2.28.1 specus hiatusque ter- 
 rae. 
 
 arbitrarius :'^^^ correct the reference from 19. 1.5 to 19.1.15. 
 
 cnmphisculus: for the formation, see Edmund Hauler in 
 Archiv, 5. 294, and Cooper, 191-192. The latter holds that 
 the form is plebeian, not archaic. For the prevalence of di- 
 minutives, especially without diminutive force, in the archaising 
 period, see Schmalz, 674 ( § 68 ), and Cooper, 185-186. In 
 
 ""R.R. 3.5.14. For Varro's archaizing tendencies see above, note 64. 
 
 '^"So blieb das vulgarlatein vielfach hinter den fortschritten der 
 gebildeten zuriick, aber es bewahrte sich auch einen grosseren reichthum 
 an formen und vvorten. . . Darum ist das vulgare oft mit dem archai- 
 schen identisch, obschon weder alles archaische vulgar ist . . . noch 
 alles vulgare archaisch zu sein braucht". G. Landgraf, Phil. Anz., 
 15.608, says: "Archaismen aber sind in der regel identisch mit vulgiir- 
 ismen". See also Cooper, xxi ; above, pages 9-10. 
 
 "'For the archaic character of adjectives in -arius see Cooper, 151- 
 155- 
 
34 Archaism in Aulus Gellius 
 
 this connection, we may note the following combinations: 
 avicula . . . parva, 2.29.3; porculis . . . minusculis, 4.1 1.6; 
 funiculo hrevi, 5.3.4; tenuissimis minutisque ossiculis, 7.1. 10; 
 surculi . . . ohlonguli, i'j.(^.'j; herediolum tenue, 19.7.1. A 
 wholly illogical combination is largiores laetioresque in con- 
 viviis invitatiunculas vi)ii.,^^- 15.2. in lemm. Compare in gen- 
 eral Gellius's handling of frequentatives : see on adiutare, 
 below, page 35. 
 
 cuius, a, um: this word is characterized as an archaism by 
 Servius on Verg. B. 3.1.^'^^ 
 
 lepidus: the citations given under this word may be greatly 
 supplemented. Add, for the positive, 10.19.2, 12.6.1, i7.i4-3» 
 18.1.12; for the superlative, 7.9. in lemm., 13.11.1, 19.4.1, 19.9. 
 in lemm., 19.9.5, ip.H-i; for lepide add 7.8.3, 9.3.2, 11.11.4, 
 13.5. 10, 13. 10.3, 18.2.6. The statement that the comparative 
 lepidior occurs but once in Gellius and once in Plautus is not 
 correct. See Cell. 17.14. in lemm. The forms of this adjective 
 and adverb occur at least twenty-five times in Gellius. It 
 should be noted also that illepidus occurs in 11.7.1, 18.4.10, 
 19.9.7; and illepidc in 11. 16.5, 16.12. in lemm., and 18.13.5. 
 These two words are doubtless to be regarded as archaisms. 
 Both are used by Plautus. The former is cited from no 
 author later than Cicero until we come to Gellius, the latter 
 but once between Cicero and Gellius, and then from the elder 
 Pliny. 
 
 necessum: as a help toward understanding the composite 
 character of Gellius's style,^^* it might have been noted here 
 that necesse is common in Gellius.^^^ 
 
 plerique omncs:'^^^ add 14.3. i pleraque omnia; 15. 7.1 pleris- 
 que omnibus; 17.5.4 plerique omnes. Plerique omnes occurs 
 also in Pronto, 183 (Naber), in a short but, from our present 
 point of view, very interesting sentence, since within seven 
 
 "* Cooper, 186-187. On double diminutives see e.g. Abbott, Repetition, 
 80-82. 
 
 ^''See Steele, A.J.P., 15.183. 
 
 ^'^T.A.P.A., 25.6, note i. 
 
 "'Cf. 5.I.3, 5-8.IO, 5.11.7, S.I3-I, 5-18.5, 7-2.5. 16.8.8. 
 
 "'Wolfflin, Lat. u. rom. Comparation, 41. 
 
Archaism in Aulus Gcllius 35 
 
 words we have two archaisms : Plcriqnc omnes qui earn cu- 
 raverant frustra fucrnnt.^''''' 
 
 stcrilus, a, um is found again in 17.21.44. In 6.1.2, how- 
 ever, we have matron- cius din stcrilcni cxistiniata>n iradunt. 
 
 ad in tare: on frequentative verbs in general see Cooper, 
 210 ff., and Schmalz. 633 (§35). It is remarked in Drisler 
 Studies, 161. that in 1.3. 13 GelHus, in paraphrasing a passage 
 from Cicero, substitutes adintarc for Cicero's adinvarc. So 
 Quadrigarius, a favorite of GelHus, uses the phrase lingnatii 
 exertarc (19. 13. 12) in describing Manlius's fight with the 
 Gaul, whereas Livy 7.5.10, in his account of the same duel, 
 writes cxscrcrc lingnam. Additional light on the complete- 
 ness of Gellius's failure to preserve in practice the true force 
 of frequentatives may be gained from a study of the follow- 
 ing passages: 3.13.1 ventitarc . . . solitnm; iyAg.2 solitnni 
 dictitare ; 18.2. in lemm. agitare soliti sijnns; 6.1.6 solitainsse 
 . . . ventitarc; 1.26.7 saef^e . . . disscrtainssc; 19.5.4 adsidne 
 dictitabat; 20.8.1 agitare erat solitus; 10.8.3 i<Jem factitatnm 
 esse credo per consuetudinem. Compare in general Gellius's 
 handling of diminutives : see above, page 33, on complnscu- 
 1ns. 
 
 coniinoliri: for verbs compounded with con- see Cooper, 262 
 ff., especially 265. Gellius borrowed from early writers con- 
 digniis, condigne, commoliri, coniplacerc, consilescerc. Com- 
 murniuratio, condecore, confabricari, conflacccscere, congermi- 
 nare, and contcmporancns are found only in Gellius, and there 
 but once. Several compounds with con- {compavcscere, con- 
 gelascere, consarcinari, convallare, convelare, convexare, con- 
 vexio) appear for the first time in his pages. '^^ 
 
 indipisci: add 12.1.24 and Brix-Niemeyer' and Wagner on 
 Plant. Trin. 224. 
 
 odi, odisse: add a reference to Georges, s.v., and Steele, 
 A.J.P., 15.188. 
 
 ad atnussiui: see Xeue-Wagener. 2.670. 
 
 adprimc: see Wofflin. Lat.u.rom. Comp.. 17-18. 
 
 '" See below, page 43. 
 
 "* For other unusual prepositional compounds in Gcllius, see T.A.P.A., 
 2513- 
 
3t> Archaism in Aldus Gcllius 
 
 ampliter: refer to Brix-Niemeyer^' and Wagner on Plaut. 
 Trin. 1060; Sonnenschein on Plaut. Rud. 265; Cooper, 200. 
 
 ast: see Schmalz, 500 (§ 252). 
 
 clam (as prep, with accus.) : see Holtze, 1.2 13; Lindsay, 
 The Syntax of Plautus, 85 ; Wolfflin in Archiv, 7.278 ; Steele, 
 A.J.P., 15.187. 
 
 compluriens: note that this form appears in Gellius 1 7.2.21, 
 as well as in 6.3.5, 'the single passage cited in Drisler Studies. 
 Hence correct the statement made here and in A.J. P., 14.218, 
 that 6.3.5 "is the only passage in Latin in which the word has 
 been preserved naturally, so to speak". 
 
 cumprimis: see Wolfflin, Archiv, 1.97; Lat.u.rom. Comp., 
 18,25 ; above, page 29. 
 
 fortassean: correct reference from 5. 14.13 to 5.14.3; and 
 see Neue-Wagener, 2.606, for full list of citations. 
 
 impendio: add 6.1.5 ^"^ 18. 12.2. In 19.7. 10 it is said of the 
 poet Laevius, item fiere . . . impendio infit, id est 'fieri im- 
 pense incipit'. Compare Wolfflin, Lat.u.rom. Comp., 20. It 
 may be noted, too, that Gellius uses impense three times with 
 an adjective, in the phrase impense doctns: 10.24.10; 13. 10.4; 
 19.7. 1. The only other example of this use seems to be in 
 Plaut. Epid. 566 impense improhiis (neither Naudet nor Gray 
 makes any comment on the word). Gellius uses impense with 
 a verb in 9.9.15, 10.3.13, 11. 18. 18, and 17.10.7. But this is 
 common, being found frequently in the first century of the 
 empire. 
 
 inibi: the list of references is incomplete. For the phrase 
 atque inibi, add the lemmata of 13.25, 14.6, 15.7, 15.27, 16.13, 
 17.16, and 18.2. This phrase does not occur in 14.7.9 or in 
 13.23.15 (cited on page 169). In the latter we have inibi 
 auteni, and in the former deinde inibi. Deinde inibi, however, 
 is practically equivalent to atque inibi}^^ 
 
 nimis quam: add Pronto 75 (Naber), and see Wolfflin, 
 Lat.u.rom. Comp., 27. 
 
 numero: see Neue-Wagener, 2.601, where Hertz's reading 
 in 20.1.54 is accepted as genuine. Hosius also reads numero 
 
 "' See Neue-Wagener, 2.658. 
 
Archaism i)i Auliis Gcllius 37 
 
 against Knapp's conjecture saepcnumero}*'* In support of 
 the accepted reading may be noted that several undoubted 
 archaisms occur in this chapter: ne . . . qiioque, 15/" 
 quitast, 52. 
 
 pone: it is interesting to compare Gellius's words in i.ii.ii 
 qui pone cum loqiientem staret, with Cicero's in 1.11.16 
 serviim . . . qui staret occultc post ipsum cum contionare- 
 tur."- 
 
 C. Archaisms of J'ocabulary Not Previously Noted. 
 
 We now proceed to the consideration of some archaisms of 
 vocabulary not included in the lists of the Drisler Studies 
 (147-171). 
 
 adfinitas meaning 'relationship by marriage' is classical. 
 With the figurative meaning, however, it is rare and possibly 
 archaic: see 1.18.5. 4-I34. 7-i-i3- Georges and Lewis and 
 Short, in addition to the references just given, cite only Varro, 
 R.R. 1. 16 and Quintilian, 1.6.24. Add, from the Thesaurus, 
 Phaedrus, 4.14.2. 
 
 aedntumus, 6.1.6 aeditumosque eius templi. In 12.10.1 Gel- 
 lius himself tells us that in his time the word was obsolete or 
 obsolescent: Aeditumus verbum Latinum est et vetus. . . 
 Sed pro eo a plerisque nunc aedituus dicitur nova et com- 
 menticia usurpatione, quasi a tuendis aedibus appellatus. 
 Evidently GeUius had seen or heard the Latinity of aeditumus 
 questioned. \'arro has the word: R.R. 1.2; 1.6.9. Cicero uses 
 it once, in Topica 8.36. 
 
 caldor, 17.8. 10 respondi . . . vinum idcirco minus cito con- 
 gelascere, quod semina quaedam caldoris in sese haberet ; 
 19.4.4 caldoremque omnem de summa corporis cute cogat ; 
 19.4.5. The word is cited three times from Varro and once 
 from Arnobius. It may be noted that in 17.8. one of the two 
 chapters in which caldor is found, two archaisms occur to- 
 gether in a single sentence: 17.8.8 Verbero,^*^ inquit ridens 
 
 '"To what Professor Knapp wrote on numcro here in Drisler 
 Studies, 170, may be added his treatment of this word in his discussion 
 of Plautus, .\mphitruo 180, in The Classical Review, 7.21-22. 
 
 "' See above, page 29. 
 
 *"Cf. the note on adiutare (ad init.) in Drisler Studies. 161. 
 
 '**See below, page 39. 
 
38 Archaism in Aulus Gellius 
 
 Taurus, nonne is curriculo'^*^^ atque oleum petis? In the same 
 paragraph we have fervit^=fervet (see above, page 32). The 
 language of the entire chapter is unusual. Calorificus, frigori- 
 ficus, incongelabilis are found only here and congelascere is of 
 Gellius's own coinage (see on comnwliri, above, page 35). 
 
 censio, 10.28.2 ex ista censione Servi Tulli ; 16.10.13 cum 
 iuventutis inopia esset, in militiam tumultuariam legebantur 
 . . . et non capitis censione, sed ... a munere officioque 
 prolis edendae appellati sunt. Evidently, here we have to do 
 with an old technical phrase."^ For Gellius's legal studies and 
 their relation to his archaizing tendencies see above, page 16. 
 The word occurs twice in Plautus, both times in a jest, but 
 once with force entirely parallel to that seen in Gellius. 
 
 fretiis, as a masculine, fourth declension by-form, is found 
 in 10.26.6 brevitas tam angusti fretus, qui terram Africam 
 Hispaniamque interfluit. In 13.21, Gellius seeks to show 
 (see the lemma) that the better writers paid more regard to 
 euphony than they did to grammatical rules. He cites various' 
 passages to prove his point, and then says in § 15: Sicuti 
 Marco etiam Ciceroni mollius teretiusque visum, in quinta in 
 Verrem frctu scribere quam freto; perangusto, inquit, fretu 
 divisa. Erat enim crassius iam vetustiusque, perangusto freto 
 dicere. Though Gellius makes fretu the later form, and char- 
 acterizes freto as obsolete already in Cicero's day, the lexicons 
 cite only two other examples of frctu from Cicero. They 
 give numerous others, however, from earlier writers, En- 
 nius, Naevius, Pacuvius, Lucilius, Varro, Lucretius, Porcius 
 Licinius. On the whole, then, fretu in Gellius may be counted 
 an archaism (especially if we remember that to him Cicero too 
 was vetus scriptor: see above, note 64). 
 
 fiagitator, 17.6. 10 eam pecuniam cum viro forte irata repe- 
 tere instituit, adponit ei flagitatorem. For this sense of the 
 word cf. Plant. Most. 768; Cic. Brut. 5.18. With two pas- 
 sages from Livy, the citations for the word cease. Apparently 
 it was obsolete in Gellius's time. 
 
 gaulus, 10.25.5, is one of a list^*'' of names of vessels which 
 
 '"See below, page 43. "'See Festus (Miill.), 65. 
 
 "" See Drisler Studies on lorea, 152. 
 
Archaism in Anlns Gellius 39 
 
 Gellius on a certain occasion remembered as occurring in the 
 veterum lihri. The word seems to be found elsewhere only 
 in Plant. Rud. 13 19. It is defined by Festus, 96. 
 
 halophanta, 8.10. lemm., is a Greek word. However, Plant. 
 Cure. 4.1.2 is the only citation given before Gellius. Cf. 
 Non. 120.8 (Lindsay), and Fest. (Miiller), loi. 
 
 halucinatio, 8.3. lemm. Non. 121.20 states that the word was 
 used by the vctcrcs. See Sen. Vit. Beat. 26.6; Arn. 4.36, 6.8; 
 Nettleship, Contributions to Latin Lexicography, 144. 
 
 hariolatio, 15. 18.3. Lewis and Short do not cite this occur- 
 rence but mark the word as found only in Cic. Div. 1.3 1.66, in 
 a quotation from Enn. (Frag. Trag. 42 Rib.). Georges gives 
 only the two instances here recorded. 
 
 libentia, 15.2.7 cui libentiae gratiaeque omnes . . . incogni- 
 tae sint. The word occurs elsewhere only in Plautus."'' 
 
 pracfica, 18.7.3 vos philosophi mera estis, ut M. Cato (fr. 
 inc. 19 J.) ait, mortualia ; glosaria namque coliigitis et lexidia, 
 res taetras et inanes et frivolas tamquam mulierum voces prae- 
 ficarum,"* 
 
 sihonis, 10.25.2. Lewis and Short give this as the only oc- 
 currence. But see Ennius Ann. 504 (Vahlen). 
 
 vasurn, as a by- form of vas, occurs in 3.14.5 pars, quae 
 deest ei va^so, though in the same paragraph vas stands as an 
 accusative in a passage quoted from Ennius. ^^^ For a full list 
 of examples, see Georges, s.v. vas. 
 
 vitor, 12.3.4; Plant. Rud. 4.3.51; Donatus ad Ter. Eun. 
 4.4.21; Arn. 2.38; Dig. 9.2.27 fin. 
 
 •verbero, a term of abuse quoted by Lewis and Short only 
 from Plautus and Terence, is found twice in Gellius. 1.26.8 
 Quid autem, verbero. nunc ego tibi ira.sci videor? and 17.8.8 
 Verbero . . . nonne is curriculo atcjue oleum petis? For 
 other archaisms in the latter chapter, see on caldor above 
 (page 37). 
 
 "'Cf. Stich. 276; Asin. 268 (name of goddess). It is read also by 
 Lorenz in Pseud. 381 (=3Q6 Goetz-Schoell) : see his note. 
 
 ""Cf. Varro. L.L. 7.70; Varro. Lucilius and Plant, ap. Xon. 66; 
 Plant. True. 495. 
 
 "•Vahlcn, .\nn. 536. 
 
40 Archaism in Aulus Gellius 
 
 illepidus: see above on lepidus (page 34). 
 
 medicinus seems to stand twice as an adjective, 17. 16.2 
 Mitridatem ilium Ponti regem medicinae rei . . . sollertem 
 fuisse; 18. 10.8 quantum habui temporis subsicivi, medicinae 
 quoque disciplinae libros attigi. Medicina ars is cited from 
 Varro L.L. 5.93, and the adjective is said to occur also in 
 Hyginus and in Augustinus.^^° With the combination medi- 
 cinae rei cf. res uxoria =^ matrimoninm 1.6.3; 4-3- ^^ lemm. 
 res cthica 1.2.4, and res cibaria 6.1.8. The w^ord early, through 
 ellipsis of ars, officina, or the like, became a noun, and as noun 
 is common and classical. 
 
 nihili, as an indeclinable adjective, is found in 2.14.2 tam- 
 quam stitisses vanum et nihili verbum esset; 10. 19.2 Homo 
 stulte et nihili; 1^.2.2 erat autem nihili homo et nugator. In 
 paragraph one of the latter chapter, we have the double ar- 
 chaism viderier gestibat (see above, pages 31-32). For nihili 
 as a virtual adjective, cf. Plaut. Cas. 245 Unde is, nihili?; Mil. 
 180; Asin. 472, 859; Bacch. 904; Pseud. 1086 (Lorenz: see 
 his note); Rud. 920; Varro L.L. 10.81; Paul, ex Fest. 175. 
 The editors of Plautus (e. g. Gray on As. 472, Brix-Niemeyer^ 
 on Mil. 180, Sonnenschein on Rud. 920) do not adequately 
 give the range of the word even in Plautus ; they afford no 
 hint of its recurrence in Gellius. In fact no editor, save 
 Lorenz, has given any serious attention to the resemblances 
 between Plautine Latin and the Latin of the archaists (and, 
 one may add, Arnobius). 
 
 percitus, as an adjective, occurs in 2.12.4 populum percitum 
 et amentem; as participle it appears in 1 5.31.3 ira percitus. 
 In its finite form the verb is cited only from Plautus (once), 
 and Lucretius (three times at least), Terence, Sallust, Cicero, 
 Livy, and then not again until Gellius. Compounds with 
 per- belong especially to the older periods of the language.^^^ 
 
 pensus, the participle of pendo, occurs as an adjective,^: 
 cams, or the like, apparently only in Gellius, 12.5.7 carius 
 pensiusqiie, and Plautus Stich. 118 utra sit condicio pensior, 
 virginemne an viduam habere. 
 
 "" See Lewis and Short. 
 "' Cooper, 284. 
 
Archaism in .lulus Gcllius 41 
 
 quercerus, 20.1.26 an tu forte morbum appellari hie putas 
 aegrotationem gravem cum febri rapida et quercera . . . ? 
 Quercera tussis occurs in a fragment of Plautus, cited by 
 Festus (Priscian gives in the same fragment quercera feh- 
 ris)}^- Quercera febris is quoted by Festus 256, from Luci- 
 lius. In Apuleius, Apologia 35, Hildebrand and Helm read 
 quercerum as a noun. See Hildebrand's note. Arnobius, 
 1.28, uses quercera as a noun. 
 
 adbibo, 2.22.25 ("o^ in Cic). Cf. Plant. Stich. 2.2.58; 
 Ter. Heaut. 2.1.8. In figurative sense it occurs in Plant. Mil. 
 3.3.10; As. 3.3.49; Ovid Tr. 3.5.14; Horace Ep. 1.2.67. 
 
 claudere, 'to halt', 'to limp', appears in 1.7.20 At si explicuit 
 diceret, inperfecto et debili numero verborum sonus clauderet; 
 4.7.4 numerus clausurus est; 13. 21. 10 sentias suavitatem soni- 
 tus claudere. In no place, it will be noted, can the conjuga- 
 tion be determined. ^^'^ The verb occurs also in Caecilius, 
 Cicero (who also has claudico), Livy (see Weissenborn on 
 22.39.3), and Apuleius (see Hildebrand on Flor. iv: Vol. 
 2.84). 
 
 deiurare: compare deiuraret 1.3.20 ; deiurasset 4.20.9; deiu- 
 rant, 11. 6.1 (note simple iuraverint in lemm.) ; deiurare, 1 1.6.5. 
 In 6. 18. 10 deieraverant occurs. According to references given 
 both in Lewis and Short ancl in Georges the word is archaic, 
 found only in ante- and post-classical writers. But neither 
 Georges nor Lewis and Short treat this word satisfactorily. 
 The former omits some of the passages cited above, and the 
 latter states, quite erroneously, that the form deiuro is now 
 retained only in Gellius 1.3.20; 1 1.6.1. For the forms of 
 peiuro, obiuro and ad-iuro in Plautus. with -n-, as well as a dis- 
 cussion of the etymology, see Minton Warren, T.A.P.A., 
 32. 110-114. 
 
 exanclare, 12.5. 10. This is Hosius's conjecture. Hertz 
 
 "" See Goetz-Schoell, Fragmenta, vs. 80. 
 
 '""Cf. Vogel, 23: "Verisimile est, Gellium, si quidem e Terentii (Eun. 
 vs. 164"), Sallustii (hist. III. 83. p. 191 ; 98. p. 300 Kr.) ; Frontonis (p. 122 
 Nieb.), Ausonii deiiique et Sjmmachi exetnplis coiijecturam faccre licet, 
 ea verbi forma usum esse, quae est claiido, dc qua vid. Priscian x. 22. 
 p. 514. Antiquior nimirum erat et minus pervolgata, quam claudeo, 
 claudico". 
 
42 Archaism in Aulus Gellius 
 
 read eluctari. Exanclare is an archaic word. For citations^ 
 see Lewis and Short. 
 
 gestito, 9.6.3. The following are the only citations in Lewis 
 and Short: Plant. Cist. 4.2.83 f . ; Cure. 5.2.4; Mil. 1.1.7; 
 Poen. 1.2. 186; Enn. ap. Cell. 1. 17.10 ( = Scenica 302, Vahlen). 
 Georges ascribes the word also to Arnob., Treb. Poll., and 
 Solin. 
 
 miiginari, 5.16.5 Sed hie aeque non diutius muginandum 
 (for the archaic syntax, see above, page 27; below, 52), 
 For the verb, cf. Lucil. and Att. ap. Non. 139; Cic. ad Att., 
 16.12. 1 ; Paul, ex Fest., 147. 
 
 pudeo, 5.1.3, is fully personal.^^* 
 
 niminari, with the meaning 'to think', 'to ponder', occurs in 
 19.7.2 figuras habitusque verborum nove aut insigniter dic- 
 torum in Laeviano illo carmine ruminabamur. It seems to 
 have this meaning only in Liv. Andron. ap. Non. 166; Varro 
 ap. Non. ibid. ; and in Varro again ap. Non. 480. 
 
 vieo, 12.3.4 a viendo 'vitor' <dictus est>. Gellius is here 
 giving examples to illustrate the fact that the root vowel of a 
 derivative does not always retain its original quantity. As 
 a verb (i.e. in forms distinct from the participle-adjective 
 vietus) the word is cited elsewhere only from the ante-classi- 
 cal period. Cf. Lewis and Short, and Georges, s.v. ; Vahlen, 
 Varia, 25, and the citations there given; Lucr. 2.1171. 
 
 admodum quam seems to be used only by Gellius 19.9. 10 
 voce admodum quam suavi versus cecinit, and Plaut. Amph. 
 541 Ex amore hie a.q. saevos est (see Ussing and Palmer ad 
 loc. Naudet made no comment). The latter passage is not 
 cited by Wolfflin, Lat.u.rom. Comp., 28. 
 
 casce, in i.io. in lemm., casce nimis et prisce loquentem, is 
 uTrai €lpiqix€vov, but it is worth while to note that its adjective 
 cascus belongs chiefly to early Latin. Cf. Ennius Ann. 24, 
 in Vahlen's edition, with Vahlen's note. An interesting side- 
 light is thrown on Gellius's vocabulary by his use of an archaic 
 stem in the very chapter in which he tells how Favorinus re- 
 bukes a young man for using old-fashioned words. This is an 
 illustration of the difference between Gellius's theory and his 
 ^^ See Knapp in A. J. P., 16.63. 
 
Archaism in Aulus Gelliiis 43 
 
 practice/''^ In the same chapter, Gellius uses abhinc with 
 the ablative.^^*^ 
 
 commodum appears as an adverb of time in 2.2.2 Taurus 
 sectatoribus commodum dimissis sedebat pro cubicuh sui fori- 
 bus. This usage is ahiiost entirely confined to Plautus, Ter- 
 ence, and Cicero's letters. ^^^ 
 
 curriculo is a mere adverb, = cito, in 17.8.8 \^erbero, nonne 
 is curriculo atque oleum petis ?^®* Note the other archaism 
 {vcrbero: see above, page 39) in this short sentence. 
 
 eadem, sc. opera, is found in 10. 1.3 as an adverb of time = 
 eodem tempore, or even as equivalent to a simple 'likewise' : 
 Is ad me rescripsit petivitque, ut rationem dicerem, cur 
 'tertium' ac non 'tertio' scripsissem. Id etiam adscripsit, ut 
 eadem quid super illo quoque mihi videretur facerem se certi- 
 orem. This use seems confined elsewhere to Plautus and 
 Terence: Capt. 293; Merc. 802; Mil. 303; Poen. 617; Pseud. 
 333; Trin. 581; Heaut. 368; Bacch. 49 eadem biberis, eadem 
 dedero tibi ubi biberis savium. In the same sense eadem 
 opera is found in Bacch. 60; Capt. 450; Most. 1039.^^^ Lind- 
 say, in his note on Capt. 459 (editio maior), remarks that 
 Plautus uses eodem and eadem opera of future actions only ; 
 to one instance, Pers. 444-445, abi . . . ad forum: eadem istaec 
 facito mulier ad me transeat per hortum, our Gellius passage 
 comes close, in logical implication, if not in form (since the 
 clause preceding that in which eadem occurs contains a virtual 
 imperative). 
 
 "* Drisler Studies, 146; above, page 27. 
 
 '^' See Hertz, Vindiciae Gellianae Alterae, 31, for explanation of this 
 usage ; below, page 49. 
 
 '" See Neue-Wagener, 2.580. For commodum, or commodo in Apulei- 
 us, see Hildebrand on De Dogm. Plat. chap, i, vol. 2.174. For the re- 
 semblances between the language of Cicero's letters and that of Plautus 
 see above, note 22. 
 
 "' See Lorenz on Plant. Most. ^62 ; Brix-Niemeyer' on Mil. 523 ; 
 Holtze. 1. 172; Neue-VVagener, 2.599. 
 
 "^* The list of references given in X'eue-Wagencr, 2.634, is not very 
 serviceable in this connection, since no distinction is made between 
 passages in which eadem is to be rendered as above, and those passages 
 in which the meaning is purely physical, and the ellipsis is of via or 
 f>arte. Xo reference is made to the passage in Gellius. 
 
44 Archaism in Aulus Gcllius 
 
 frustra esse^^^ is said of persons in 5.1.1 'Cum philosophus\ 
 inquit, 'hortatur, . . . turn, qui audiunt, si . . . obvias vulgatas- 
 que laudes effutiunt . . . turn scias et qui dicit et qui audiunt 
 frustra esse'. The usage is found several times in Plautus, 
 especially in the phrase ne frustra sis. See Lorenz on Mil. 
 1422 and on Most. 567; Ennius Sat. 62 (Vahlen) Qui frus- 
 tratur is frustra est, si non ille est frustra. The phrase oc- 
 curs in this use, be it noted, in Sallust (lug. 85), whose archa- 
 izing tendencies are well known. Cf. also Apul. De Mag. 19 
 frustra es, and Pronto 183.9 (Naber) Plerique omnes . . . 
 frustra fuerunt (here two archaisms occur together: see above, 
 page 34, on plerique omnes). Lorenz on Mil. I.e. wrongly 
 refers to Fronto 189.3. Note, finally, that in 5.1.1 we have 
 another archaism, illi = illic.'^^'^ 
 
 multuin, as a strengthening adverb with an adjective, ap- 
 pears in 13.20.8 ipse quoque iam multum senex. The usage 
 is archaic and plebeian. ^'^- 
 
 nimio, as a strengthening adverb with adjectives, occurs in 
 1.3.25 cum vero amici utilitas nimio est amplior; 10.26.9 
 nimio confidentius ; 19.5.3 aquam nivalem . . . hominibus 
 potu nimio insalubrem (twice, it may be noted, with a com- 
 parative, once with a positive). As examples of nimio with 
 a positive Wolfiflin^^^ cites Plant. Bacch. 396, 770; True. 4.1.6; 
 Stich. 704; Naev. (vs. 13 in Ribbeck's Com. Rom. Frag). 
 But Ribbeck^ writes nimium, and Langen, Beit. z. Krit. u. 
 Erklar. des Plautus, 333-335, maintains that riiniio occurs in 
 Plautus only with a comparative; with the positive he would 
 read nimium, though to do this he must emend in at least one 
 place. Lorenz, in his note on nimio celerius, Most. 72, does 
 not discuss this point. In Bacch. 770 Lindsa}^ still reads 
 
 ""See Wolfflin, Archiv, 2.1; Nagelsbach, Lateinische St^listik^ 577. 
 
 "'Drisler Studies, 169. 
 
 ^^ See Lindsay (editio maior) on Capt. 272 non multum fuit molesta 
 servitus; Schmalz, 613 (§ 9, Anm. i) ; Wolfflin, Lat. u. rom. Comp., 
 8-9; Neue-Wagener, 2.581. According to Wolfflin, this usage is found at 
 least ten times in Plautus, never in Terence, four times in Horace's 
 Satires and Epistles, and then chiefly in " archaisirende oder vulgare 
 Auctoren". 
 
 ^^ Lat. u. rom. Comp., 24. 
 
Archaism in /lulus GcUius 45 
 
 Nimio illaec res et magiiae (MSS mane) dividiae mihi 
 though Langen would emend nimio to nimis and mane to sane 
 (Leo reads nimium). So in True. 704 Lindsay reads tum 
 illuc nimio magnae melHnae mihi ; Langen emends to nimiom 
 (nimiujn), and Leo accepts this reading. In Bacch. 396 nimio 
 impendiosum praestat te quam ingratum dicier (which equals, 
 in sense, nimio melius est, etc., or nimio satius est, etc.), and 
 Stich. 699-700 (not cited by Wolfflin) Utrum Pontine an 
 Libero imperium te inhibere mavis? SA. Nimio liquido Li- 
 bero, a comparative is clearly present in sense. On the basis, 
 then, of Lindsay's^^* excellent text of Plautus we have to say 
 that Plautus used nimio with both positives and comparatives, 
 though more frequently with the latter; Gellius's use of the 
 word, then, is in exact agreement with Plautus's. For nimio 
 with a comparative, see Most. y2, 442; Bacch. 151. nimio plus, 
 found e.g. Bacch. 122, Most. 1103, is not uncommon: see Lucr. 
 5.988; Horace C. 1. 18.15; 1.33.1; Epp. 1. 10.30; Livy (see 
 Weissenborn on 1.2.3). Hor. Epp. 2.1. 198 has ni]nio plura. 
 Cf. also Antonius ap. Cic. ad Att. 10.8A.1. See Schmalz, Lat. 
 Synt., § 97. Gellius also uses nimium in this way. See 
 17.21.47 non nimium longe ; 19.7.13 nimium poetica; 20.1.27 
 morbus . . . non febriculosus neque nimium gravis. This is 
 especially characteristic of Plautus. ^"'^^ 
 
 oppido'^^''' is forcibly characterized by Quint. 8.3.25 as obso- 
 lete, yet it occurs in Gellius 13.5.9 oppido bonum (sc. vinum). 
 oppido quam is found in 2.23.3 oppido quam iacere atque 
 sordere incipiunt ; 15. 30.1 oppido quam . . . inepti et f rivoli ; 
 16.7. 1 oppido quam verba finxit praelicenter ; 17.12. i oppido 
 quam libens. 
 
 prorsus retains its original physical meaning in 2.30.4 venti 
 a septentrionibus ex altiore caeli parte in mare incidentes 
 
 '"Lindsay, Sjntax of Plautus, 80 (§ 3), discusses nimio very inade- 
 quately; he remarks merely "normally with Comparative, but cf. Bacch. 
 770, True. 704, Naevius com. 13". 
 
 "'Langen, I. c, gives over thirty examples. See Gildersleeve- 
 Lodge, 439, N.3. 
 
 "•See Schmalz, 613 (§ 9. Anm. i), page 665 (§ 37, ad fin. : the section 
 is labelled ".Vrchaismen") : Xeue-Wagener, 2.602. Schmalz notes that 
 oppido occurs in .Vpuleius, but makes no reference to Gellius. 
 
46 Archaism in Aulus Gellius 
 
 deorsum in aquarum profunda quasi praecipites deferuntur 
 undasque faciunt non prorsus impulsas, sed iniitus commo- 
 tas. The meaning is fixed by § 5 Austri vero . . . inferiores 
 et humiles per suprema aequoris euntes protrudunt magis 
 fluctus quam eruunt. Cf. 16.19.17 Turn Arionem prorsus ex 
 eo loco Corinthum petivisse. See Hertz, Vindiciae Gellianae 
 Alterae, 60. In 9.4.6 Hertz reads, though doubtfully, vestigia 
 pedum habentes retro porrecta, non ut ceterorum hominum, 
 prosum spectantia ; so Hosius. Both Hertz and Hosius follow 
 the conjecture of H. Hagen, in Bursian's Jahresber., 1873, 
 1415; the MSS are corrupt. For other examples, see Cato ap. 
 Fest. 234; and Neue-Wagener, 2.746. In Plant. Pseud. 955 
 the Palatine text gives non prorsus verum ex transversa cedit. 
 The Ambrosian palimpsest, according to Studemund's Apo- 
 graphon, clearly shows NONPROSUSUE. But editors (e.g. 
 Lindsay, Morris, Lorenz, Leo) give the verse as it is cited 
 in Varro L.L. 7.81 Ut transvorsus, non provorsus cedit, quasi 
 cancer solet.^*'^ provorsus, if read, is, of course, the original 
 form of prorsus, prosus, prorsum, prosum. 
 
 rarenter,'^^^ 3-i6.i ; 17.8.9. In 10.15.4 the word is probably 
 to be ascribed to Fabius Pictor. 
 
 tractim,^^^ 4.6.6 littera <i> scilicet tractim pronuntiata; 
 6. 10. 1 ; 6.20.3 vocalis . . . tractim sonat. 
 
 Without doubt Professor Knapp is correct in his statement^^° 
 that Gellius's favorite authors were Plautus, Ennius, and Cato. 
 We should, then, expect Gellius to borrow many words from 
 his favorites. Even a casual examination of the word lists in 
 the Drisler Studies and in this paper will prove that such is 
 the case. 
 
 The following words or uses of words seem to be taken 
 from Plautus :^^^ ambulacrum, fictura, fidicina, ingratis, porcu- 
 
 ^" One wonders why the editors forget that the ancients do not always 
 cite accurately; see e. g. Jebb on Sophocles, Antigone, 223, 909 ff. 
 
 "*See Neue-Wagener, 2.735. 
 
 "' See Neue-Wagener, 2.567, where it is cited from Plautus, Ennius, 
 Lucretius, Vergil, Sempronius Asellio. 
 
 "" Drisler Studies, 132-134. 
 
 "'See T. A. P. A., 28 (1897)- vi. 
 
Archaism in Aldus Gellvus 47 
 
 lus, saviatio, scitamenta, symbola, amasius, condignus, crucia- 
 bilis, cuias, exoticus, lepidus, manifestarius, ridicularius, sum- 
 mas, inceptare, esitare, indipisci, percupio, protolli, perservire, 
 discussed in the Drisler Studies. From the present paper may- 
 be added curriculo, flagitator, libentia, verbero, nihili, pensus, 
 quercerus, commodum, eadem, multum, nimio. 
 
 Citations from Ennius, Cato, and Varro are extremely 
 frequent in the notes in both papers. Thus we find GelHus's 
 love of the older writers and his deliberate imitation of them 
 proven from two independent lines of research: first, a col- 
 lection of the passages in which he refers to them, always in 
 the highest terms of commendation ; and, secondly, a careful 
 study of his vocabulary. 
 
PART II 
 
 Archaisms of Syntax 
 
 Turning now to archaisms of syntax, we shall dwell first on 
 some points in Gellius's use of the cases. 
 
 In 16.19.10 parco is constructed with the accusative: Tum^ 
 ilium . . . vitam modo sibi ut parcerent oravisse. Cf. Plaut. 
 Most. 104 (a corrupt line) ; Cure. 381 ; Cato, R.R. 155. See 
 Sonnenschein- and Fay on Most. 104 and Lorenz's critical note 
 on that verse (pages 206-207); Nettleship on Aen. 10.532; 
 Schmalz, 372 (§ 84). In Apul. Met. 1.2, 1.8, we find in the 
 editions parco in and acc./^- two manuscripts, however (R, f ) 
 show parco with the accusative. See Hildebrand ad loc. 
 Servius on Aen. 10.532 says parco with accusative was used by 
 Plautus, Lucilius and Ennius ; see Conington ad loc, Steele, 
 A. J. P. 15.179. 
 
 euro takes the dative in 1 7.9.1 Libri sunt epistularum C. 
 Caesaris ad C. Oppium et Balbum Cornelium, qui rebus 
 eius absentis curabant. Gronovius ad loc. remarks on the 
 reading that it is "doctius quam ut potuerit ab librario venire". 
 Cf. Plaut. Rud. 182 si tu de illarum cenaturus vesperi's, illis 
 curandum censeo, Sceparnio ; Rud. 146 Amori haec curat, 
 tritico curat Ceres (here Professor Sonnenschein remarks, 
 "euro with dat. 'care for' is ante- and post-classical") ; Trin. 
 1057 Sed ego sum insipientior qui rebus curem publicis (see 
 Brix-Niemeyer^ ad loc); True. 137; Att. 143 Ribbeck. In 
 later Latin the dative is to be found also in Apuleius : Hilde- 
 brand, 2.1 17, on De Deo Socr. 2, cites De Deo Socr. again, 
 cap. 16; De Mag. 36; De Mundo, 30. Cf. also Macrob. 
 1. 14.6 sacerdotes, qui curabant mensibus ac diebus ; Tert. 
 Apol. 46; Pronto 228. 19 (Naber) Tum lovem ferunt . . . 
 cum suo corde agitasse de suis germanis fratribus unum 
 praeficere, qui nocti atque otio hominum curaret. See 
 Schmalz, 372 (§ 84), and, more particularly, C. F.W. Miiller, 
 Syntax des Nominativs und Akkusativs in Lateinischen (Sup- 
 
 "*The construction here is akin to that seen in Lucr. 6.399 cur ipse 
 sinit neque parcit in hostis. See Professor W. A. Merrill, ad loc. 
 
Archaism in Aldus Gclliiis 49 
 
 plement to Stolz, Historische Grammatik), 123-124. Miiller 
 notes that procurarc also is construed with the dative, e.g. 
 in Plautus, Apuleius, and the law writers (so in a senatus 
 consultum ap. Gell. 4.6.2). We may, then, refer here again 
 to Gellius's legal studies as a factor in the development of 
 his style: see above, page 38, on ccnsio; below, note 175. 
 
 capitis is used with pcrd<cre in 1.3.4 Ita lex fuit, uli eum 
 hominem condemnari necessum esset. Aut amicus igitur 
 capitis perdendus aut adhibenda fraus legi fuit. For the 
 phrase, which is evidently modelled on capitis dauinarc, cf. 
 Plaut. Asin : 132 capitis te perdam ego et filiam ; Bacch. 489 
 Egone ut illam mulierem capitis non perdam? Mil. 371 quern 
 pol ego capitis perdam. In the Brix-Niemeyer'* edition of 
 the Miles Gloriosus (1901), there is no hint that the phrase 
 recurs after Plautus. Gray, too, on As. 132 is silent on this 
 point. 
 
 intcrficcre is used with ace. and abl. in 12.7.2 Eadem mulier 
 virum ct filinm . . . venenis clam datis vita inter fecerat. Cf. 
 Plaut. True. 518 Salve, qui me interfecisti paene vita et lumine. 
 Naudet (see his note on True. 2.6.37) was aware that the con- 
 struction under discussion occurred in Gellius. 
 
 abhinc is twice coupled with the ablative, instead of with 
 the accusative; see 1.10.2 sermone abhinc multis annis iam 
 desito uteris, and 14. 1.20 centesimo usque abhinc saeculo. 
 Hertz, Vind. Gell. Alt., 31, defends this construction against 
 Madvig's criticism (expressed in his Adversaria Critica, 2). 
 Part of Hertz's answer deserves to be quoted here. Madvig 
 had said that there was no reason to suppose that Gellius had 
 departed in 1.10.2 from the "perpetua lex" according to which 
 abhinc was coupled with the ace. Hertz rejoins, inter alia, 
 thus : "Aber G. kehrt sich eben nicht an solche perpetua lex, 
 d.h. an den Sprachgebrauch der sogenannten mustergiiltigen 
 Classiker KaT*c^o;(i7v,wenn er bei den von ilim bev^orzugten 
 archaischen Schriftstellern einen abweichenden seltenen Ge- 
 brauch findet . . . Und so schloss G. sich audi hier an Plau- 
 tus . . . ". Hertz has in mind Most. 494; see Lorenz's note. 
 
50 Archaism in Aulus Gellius 
 
 For an archaism of vocabulary in this chapter see above, page 
 42, on casce. 
 
 penetrare is used with a reflexive pronoun in 5. 14. 18 specum 
 . . . nanctus remotam latebrosamque, in earn me penetro et 
 recondo, and 13.10.1 Labeo Antistius ... in grammaticam 
 sese atque dialecticam litterasque antiquiores altioresque pene- 
 traverat. The construction seems to be found elsewhere only 
 in Plautus. Brix-Niemeyer^ on Trin. 146 cite penetrare se 
 from Amph. 250, True. 44, and Trin. 276; penetrare me from 
 Trin. 291, 314; penetrare pedem from Men. 400, 816. They 
 state also that in Bacch. 66 the verb is intransitive, but Goetz- 
 Schoell read there penetrem me (after Bothe). Lindsay, with 
 extraordinarily bad punctuation, reads : Quid ego metuam, 
 rogitas, adulescens homo? penetrare [me] huius modi in 
 palaestram, ubi damnis desudascitur? The question-mark 
 after homo should be changed to a comma. Professor Ben- 
 nett's treatment of this construction in his Syntax of Early 
 Latin, Vol. I — The Verb (page 5, under § 3) is inadequate. 
 
 Gellius several times has the genitive of a noun in depend- 
 ence on the genitive of a gerund :^^^ cf. 3.16.1 Multa opinio 
 est . . . gigni hominem septimo rarenter, numquam octavo, 
 saepe nono, saepius numero decimo mense, eumque esse homi- 
 num gignendi summum finem; 4.1 5.1 Elegantia orationis Sal- 
 lustii verborumque fingendi et novandi studium cum multa 
 prorsus invidia fuit; 5.10.5 Euathlus . . . eloquentiae discen- 
 dae causarumque orandi cupiens fuit; 16.8.3 fecisse videtur 
 eum librum Aelius s^ii magis admonendi quam aliorum docendi 
 gratia; 7.14.4 non sane dignum esse imponendi poenae studio 
 visum est.^^* The construction is found in Plautus, Terence, 
 Varro, and Cicero : see Brix-Niemeyer® on Plaut. Capt. 852 ; 
 Draeger 2.831, 832; Gildersleeve-Lodge, § 428, N. i; Holtze 
 2.54; Roby, 2.1xviii. One example is cited from Pronto, 188 
 
 "^ For another view of the construction, see below, pages 51-52. 
 However the construction is to be explained, we have to do here with 
 a usage obsolete by Gellius's time. 
 
 "* Hertz defends his readings in the passages cited above in Vind. 
 Gell. Alt., 58,59 in a full discussion. Hosius, be it noted, reproduced 
 all these readings. 
 
Archaism in Aulns Gcllius 51 
 
 (Naber) : Neque enim cum alio ullo . . . mihi . . . est . . . 
 tantus usus studioniiu bonarumqiie artium communkandi. 
 
 Three things may be noted in this connection. 
 
 (i) In 5.10.5 eloquentiae discendae causarumque orandi 
 cupiens fuit, the normal and the abnormal construction appear 
 together. On such inconsistencies compare the remarks made 
 above, page 27, and Hertz, Yindiciae Gellianae Alterae, 59 : 
 "wenn aber G. diese Construction sonst ofter in normaler 
 Weise anwendet, so ist es bei ihm nicht nur ohne Anstoss, 
 sondern recht im Charakter, wenn er audi einmal eine anomale 
 und seltene Form gebraucht, sei es nach dem Vorbilde des 
 Ennius und des Cicero, sei es nach dem des Ennius allein oder 
 schliesslich .... ohne einen, wenigstens ohne einen uns be- 
 kannten Vorganger". 
 
 (2) In 7.14.4 the dependent genitive is in the singular. 
 
 (3) If we accept the theory that in the examples quoted 
 above the genitive of the noun depends on the gerund, we may 
 find a simple explanation of such a familiar idiom as that 
 seen e.g. in Cicero in Cat. 1.7 multi principes civitatis Roma 
 non tarn sni conservandi quani tuorum consiliorum reprimen- 
 dorum causa profugerunt. Sui may here well be a pronoun, 
 depending upon the gerund as objective genitive. This view 
 has occasionally been upheld, e.g. by Allen and Greenough in 
 their note on this passage, and in their Latin Grammar (504. 
 c). But other Latin Grammars in common use — e.g. Bennett, 
 339-5» Gildersleeve-Lodge, 428 R.i. — see a gerundival construc- 
 tion. Sui is in that case, of course, neuter, not masculine, 
 meaning 'their own being' or the like. 
 
 Yet another view of the construction is held by competent 
 scholars. According to Schmalz, 444, in all the expressions 
 cited above, the two genitives are equally dependent on the one 
 governing noun; so Bennett. Syntax of Early Latin. V^ol. I. — 
 The Verb, 449 (under 4). So, too, in effect Roby. in his 
 Latin Grammar (1874), 2.1xviii. This view appears need- 
 lessly complex, and seems to disregard the plain implications 
 of language, in order to find a way of escape from the ad- 
 mission of a nonclassical construction into Plautus. Surelv, in 
 this instance, one need not resort to extreme measures in the 
 
52 Archaism in Aulns Gclliiis 
 
 attempt to bring this construction into harmony with classic 
 usage, for, to remain entirely within the limits of the present 
 paper, we should still have to explain euro and parco with the 
 accusative, penetro with a reflexive pronoun, etc. (see above, 
 pages 48-50). The editors of Plautus, however, have not 
 infrequently accepted the explanation adopted in the present 
 paper: see e. g. Elmer and Hallidie on Capt. 852. Professor 
 Morris, however, in his note on the passage harks back to 
 Roby's view (so, too, Brix-Niemeyer'^') ; Lindsay ad loc. 
 (editio maior) quotes with approval part of Professor Mor- 
 ris's note, but not enough to show to one who did not have 
 Professor Morris's whole note before him what view he took 
 of the construction as a whole. See also Kiihner-, 2.744-745. 
 
 Gellius, at times, joins an accusative to a neuter gerundive 
 / coupled with est in an impersonal construction : 4.9.9 templa 
 quidem ac delubra . . . quae non volgo ac temere, sed cum 
 castitate caerimoniaque adeundum . . .; 5.16.5 Sed hie ea quae 
 non diutius muginandum (so Hertz. Hosius, however, after 
 Petschenig, reads: Sed hie aeque non diutius muginandum). 
 Hertz discusses the readings in Vind. Gell. Alt., 50-51. See 
 also Kiihner^, 2.734; Draeger, 2.821 ; Schmalz, 441. According 
 to Schmalz, the construction occurs once in Plautus (Trin. 
 
 869), "ofters bei Lucr Varro und Catul bei 
 
 Verg. und seinen Nachahmern, nirgends bei Caes., Sail., Liv., 
 
 Tac Spater lesen wir bei Juristen,^'^. . . bei Tert., Pall. 
 
 4 . . . " Gellius, it will be noted, is not mentioned. Kiihner, 
 too, fails to note that the construction occurs in Gellius. 
 
 The construction was briefly discussed long ago (1874) by 
 Roby, Latin Grammar, 2.1xxii fif. He gives a list of examples 
 meant to be exhaustive (except for Varro), but cites no in- 
 stance from Gellius. See also various editors, e.g. Brix-Nie- 
 meyer^ on Plant. Trin. 869 ; Munro and Merrill on Lucr. 
 I.I 1 1 (especially the authorities referred to by Merrill) ; Ellis 
 on Catullus, 39.9. 
 
 A most interesting passage in this connection is Cicero, Cato 
 Maior 6 quam (viam) nobis quoque ingrediundum sit. See 
 Reid ad loc. (page 75 ; also pages 27-28) ; Reisig-Haase, Vorle- 
 
 "'^ See above on censio, page 38; page 48, s.v. euro, ad fin. 
 
Archaism in Aulus GcUins 53 
 
 sungen, 3776. Reisig-Haase see here a deliberate archaism ; 
 Professor Reid, however, holds that Cicero is not imitating 
 Cato Censor, since in the extant writings of Cato the con- 
 struction does not occur. Schmalz, 441, after citing Cicero, 
 CM. 6, and Scaur. 13, as giving examples of the accusative in 
 dependence on the gerundive, says: "(Cicero) mag sich ge- 
 scheut haben, vom Deponens, dessen transitive Bedeutung sich 
 nicht fest ins Bewusstsein eingelebt hatte, die personliche 
 Konstruktion zu verwenden". 
 
 According to Schmalz, 465, the supine in -um is very com- 
 mon in Plautus, Terence, and old Latin generally, less com- 
 mon in Cicero and Caesar, more frequent again in Sallust, 
 V^arro, in the Bell. Afr. and Livy. "In der nachliv. Prosa", 
 he continues, "gehort das Supin fast nur den Archaisten und 
 den von Sail. u. Liv. abhangigen Schriftstellern an".^^" While 
 the accusative of the supine is common in Gellius, it is much 
 more frequent with an object than without. For the supine 
 without an object see 6.14.8 quos . . . legaverant impetratum ; 
 1 2. 1 3.3 issem . . . sciscitatum. The following are examples of 
 the supine with an accusative: 2.29.6 Dum . . . iret cihum 
 pullis quaesitum; 3.13.2 cognoscit currere eos auditiim Callis- 
 tratum; 6.3.7 socios . . . defensum conservatunique pergit ; 
 6.3.44 servatum ire socios niteretur; 9.15.3 It auditum (sc. 
 eum) lulianus; 10.6.2 Utinam . . . frater . . . istam mnlti- 
 tudiuevi perditum eat; 10. 19.3 qui . . . pcccatmn sitiini . . . 
 excmptum purgatnmque ibat ; 12.1.2 Eamus et piicrum visum 
 et patri gratulatum ; 12. 1.9 Quod sit . . . odio dignum . . . 
 hominem . . . interfectuni ire; 14.6. i aditituiii oniatumqwe 
 volo ire N^octcs tuas; 14.6.5 Nam meae Xoctes. quas instntctu»i 
 ornatiimqiic isti ; 16.11.6 Psyllos . . . decretumque fecisse uti 
 . . . iure belli res pctitmn proficisceretur ; 18.5.3 Eamus . . . 
 auditum. . Astum Ennianistam; 16.5.9 qui dominum. . .salu- 
 
 "'See the statistics in Draeger, 2.857-865, especially § 608; American 
 Journal of Philology. 32.29, note i, in a review of Vahlen's Ennius ; 
 Grandgent, § 103; Bennett. Syntax of Early Latin, r. § 453 ff. ; Lindsay, 
 Synta.x of Plautus. pages 76 ff. ; Gildcrslecve-Lodge. 434. note 3; 
 Frohenius, Die Syntax des Ennius ^ 137 (=page 67). While the ac- 
 cusative of the supine is frequent in .\puleius, it is significant that no 
 instance of the supine with an accusative is cited. 
 
54 Archaism in Aulus Gellius 
 
 tatum venerant. It is to be noted that the construction occurs 
 chiefly in dependence on forms of ire: see again Schmalz, 465; 
 Draeger § 609. 
 
 Twice GelHus uses opus est with the ablative of a perfect 
 passive participle: 14.2. 16 Atque illud amplius ambigi ac dubi- 
 tari scio, debeatne iudex inter cognoscendum ea, quae dicto 
 quaesitoque opus est, dicere et quaerere; 15.22.8 ait in quiete 
 cervam . . . quod opus esset facto, praedicere. The presence 
 of quae and quod in these examples makes them but the more 
 noteworthy. See Lorenz on Plaut. Most. 48 ; Dziatsko-Hauler^ 
 on Ter. Phorm. 584; Holtze, 1.139-141 ; Kiihner^ 2. § 128; 
 Draeger 2.780 ; Reisig-Haase, Vorlesungen, 392 ; Bennett, Syn- 
 tax of Early Latin, 1.439-440. 
 
 I have made an entirely independent examination of the pre- 
 positions in Gellius. On comparing my results with what had 
 already been done in this field, I found that (with possibly three 
 exceptions which are discussed below) my own investigations 
 failed to reveal any clear examples of archaistic usage that had 
 not already been considered by Professor Knapp."^ 
 
 It will be sufficient, then, in the case of the prepositions which 
 have already been considered (consult note 177, below), merely^ 
 to give a list of the phrases in which they occur and refer to 
 the page on which they are discussed. It is not intended that 
 the citations under each example shall be exhaustive. 
 
 a: 16.6.8 a te rogavi; 17.10.7 oravitque a suis amicissimis ; 
 13. 31. 10 cum valebo ab oculis (l.c.14). 
 
 absque: ^^^2.26.20 absque te . . . uno forsitan lingua profecto 
 Graeca longe anteisset; 2.2.7 absque praeiudicio. 
 
 ad: 6.17.11 revises ad me; 18.8.2 quod ad se . . . non viseret; 
 13. 31. 10 revise ad me; 19.10.1 ad Frontonem ... ire et visere 
 (1.C.6). 
 
 clani:'^''^ 2.23.16 ea res clam patrem fuit (l.c.6). 
 
 cum (a use of cum not included in Notes on Prepositions) : 
 1 8. 10.7 cum dis bene volentibus opera tua sistas hunc nobis 
 sanum atque validum quam citissime. Schmalz, § 139, says 
 
 "'Notes on Prepositions in Gellius, T. A. P. A., 25 (1894). ^-32- 
 
 "' See Drisler Studies, 167. 
 
 "'See Drisler Studies, 168; above, page 36. 
 
Archaism in Aulus Gellius 55 
 
 that this phrase occurs only in Ennius, Cato, and GelHus. The 
 passages are Ann. 201 (Vahlen) ; R. R, 141. 
 
 in: 3.1.4 in eo ipse quaerendo fui; 3. 7. 12 in exspectando 
 sunt; 1 1.5.3 ^^ quaerendo semper considerandoque sunt; 11. 
 16.6 in cogitando fuissem; 13. 31.9 in legend© ; 18.10.7 in 
 medendo quani in dicendo (I.e. 14). 
 
 in medium relinqnerc (a use of in not included in Notes on 
 Prepositions) : 7.14.9 anne autem quasi omnino parvam et con- 
 temptu dignam praeterierit poenae sumendae causam propter 
 tuendam laesi hominis auctoritatem, an magis quasi ei, quam 
 dicebat, rei non necessariam praetermiserit ... in medium 
 relinquo. In 17.2. 11 Gellius quotes from Quadrigarius the 
 words nos in medium relinquemus, with the following com- 
 ment : Vulgus in medio dicit ; nam vitium esse istuc putat. In 
 the same chapter Gellius calls nc . . . quoque and cuinprimis 
 obsolete, but he uses them both (see above, pages 29 and 
 
 prae:S-^2)-Z Unum hoc . . . quod prae manibus est, ponemus 
 (local sense) ; 19.8.6 si . . . liber prae manibus est (I.e. 29). 
 
 subter: 10.15. 15 subter arborem felicem (I.e. 31). 
 
 Passing now to the conjunctions, we note quando in temporal 
 meaning.^""^ Some fourteen passages occur in Gellius in which 
 
 '*" Similar is the phrase /;; f>otestatem esse, which an unnamed friend 
 of Gellius defends in i. 7. 17 (cf. § 4), quoting Cicero, De Imp. Cn. 
 Pompeii 33 ; he argues that this is no solecism, but that the Greeks use 
 this idiom, et Plautus verborum Latinorum elegantissimus in Amphitru- 
 one <i8o> dixit: num vero mihi in mentem fuit, non, ut dici solitum 
 est, in mente. In § 18 he adds : multam .... apud veteres scriptores 
 locutionum talium copiani offendimus atque his vulgo adnotamcntis in- 
 spersimus (a very significant passage, lighting up what was said above, 
 page 27, note loi, about Geliius's 'revival' archaisms). He explains the 
 use of the accusative in the phrase on the score of euphony (§ 19). 
 For the idiom in mentem est see Amph. 710; Bacch. 160; Ter. Ad. 528 
 (with Ashmore's note ad loc). Palmer (Amph. 710) says: "Sallust 
 has several instances of this idiom". Clearly these phrases were ob- 
 solete in Geliius's day; hence his discussion of them throws some 
 light on in inedium relinquo. 
 
 '"See Draeger, 2.582: "Als relativische Temporalpartikel kommt es 
 besonders im archaischen Latein vor und wird von da ah weniger 
 gebraucht. Es bezeichnet sehr selten eine einmalige. in der Kegel und 
 
56 Archaism in Aulns Gclliiis 
 
 quaiido seems clearly to have temporal force. Compare, for 
 examples involving the indicative, usually with iterative force, 
 1.20.3 Solidum est, quando non longitudines modo et latitudines 
 planas numeri linearum efficiunt, sed etiam . . . altitudines ('We 
 have a solid body, whenever', etc.) ; 6.17.4 remotiora ... si 
 discere et scire debuero, quando usus mihi venerit, tum quaeram 
 ex te; 7.14.4 Quando igitur spes . . . magna est (note just 
 above, in same paragraph, Tertia ratio vindicandi est, . . . cum 
 poenitio . . . necessaria est. Iterative clauses with cum and 
 the indicative are common in Gellius. Here the long quando 
 clause is resumed by quicquid ita delictum est) ; 9.9.1 Quando 
 ex poematis Graecis vertendae . . . sunt insignes sententiae, non 
 semper aiunt enitendum ; 11. 1.4 Quando igitur nunc quoque . . . 
 multa dicitur vel minima vel suprema, observari solet; 11.3.1 
 Quando . . . otium est . . . aut spatiamur aut vectamur, quaerere 
 nonnumquam aput memet ipsum soleo; 17.6.6 quando mulier 
 dotem marito dabat, tum, quae ex suis bonis retinebat neque ad 
 virum tramittebat, ea recipere dicebatur; 17.7.6 quando (sc. 
 est and erit) per sese ponuntur, habent atque retinent tempus 
 suum; 17.9.9 Quando usus venerat . . . conpHcabant ; 19.8.1 
 quando erat a magistris . . . otium, ad Frontonem . . . perge- 
 bam sermonibusque eius . . . fruebar; 19.8.15 quando forte erit 
 otium, quaerite (not necessarily iterative). 
 
 For instances involving the subjunctive see Praef. i iucundi- 
 ora alia reperiri queunt, ad hoc ut liberis quoque meis partae 
 . . . remissiones essent, quando animus . . . indulged potuisset ; 
 Praef. 2 annotabam . . . ut quando usus venisset aut . . . oblivio 
 tenuisset, et libri, . . . non adessent . . . foret (in the latter in- 
 stance, the subjunctive is clearly accounted for by oratio 
 obliqua; in the former, the subjunctive is due to attraction); 
 7- 1 3-5 Quaesitum est, quando moriens moreretur . . . et quando 
 surgens surgeret . . . et qui artem disceret, quando artifex fieret ? 
 In the case just cited there is, of course, no iterative force and 
 the question is dependent (we have oratio obliqua, then, once 
 
 in der alten Zeit immer eine wiederholte oder zu unbestimmter Zeit 
 geschehende Handlung. Der Modus ist iiberall der Indivativ". Cf. 
 also Schmalz, Lat. Synt. § 266; Gildersleeve-Lodge, § 580, Note 3; 
 Grandgent, Vulgar Latin, § 82; Lindsay, Syntax of Plautus, 135. 
 
Archaism in .lulus Gcllius 57 
 
 more). Gellius is not fond of the iterative subjunctive. He 
 uses it but once with ubi (i 7.2.1 ubi venisset usus), and seldom 
 with d'U))i or cu))i. The causal quando is not common. Gellius 
 prefers quia, quod, and especially quoniam. 
 
 Schmalz, 566 (§326), recognizes no example of quoniam 
 with temporal force outside of Plautus and Terence.^*- But it 
 certainly seems to have temporal force in Gellius 6.5.4^"'^ Polus 
 unice amatum filium morte amisit. Eum luctum quoniam satis 
 visus est eluxisse, rediit ad quaestum artis.'""* Quoniam seems 
 to be temporal, with iterative force, in 15.27.5 Propterea cen- 
 turiata in campo Martio haberi exercitumque imperari praesidii 
 causa solitum, quoniam populus esset in suflFragiis ferendis oc- 
 cupatus.^*^ 
 
 With reference to causal cjnn/®" in Draeger 2.680 we find the 
 following: "Seit der klassischen Zeit ist der Konjunktiv der 
 allein gebrauchliche Modus." Similarly Schmalz, page 565 : 
 " Das kausale, adversative und konzessive quoni wird seit der 
 klassichen Zeit ausschliesslich mit dem Konjunktiv konstru- 
 iert". 
 
 The causal force, however, seems clear in certain passages 
 from Gellius, although the nuance may be such that in some of 
 the instances the temporal force might be admitted. See 2.29.1 
 Aesopus ille e Phrygia fabulator haut inmerito sapiens existi- 
 matus est, cum, quae utilia monitu suasuque erant non severe 
 neque imperiose praccepit et censnit, ut philosophis mos est, sed 
 festivos delectabilesque apologos commentus, res salubriter ac 
 
 "^ Compare Sonnenscheiii on Rud. 67: Gray on .\sin. 350; Lorenz 
 and Brix-Xiemeyer' on Mil. 129; Brix-Xiemeyer' on Trin. 14. 
 
 "* But see Apul. Ap. § 17: ipse .... scriptum reliquit tris 
 
 servos solos ex urbe duxisse, quoniam ad villam publicam venerat. 
 
 "*Cf. Gildersleeve-Lodge, 580, X. 3. In earlier editions Schmalz too 
 saw temporal force here. 
 
 *** Compare Weiss's translation : "Deshalb pflegten die Centuriat- 
 Comitien auf Marsfelde abgehalten und das (waffenfahige") Volk zur 
 Besitzung des Wahlplatzes aufgefordert zu werden des Schutzes und 
 der Sicherheit halber, so lange als das \'olk beim Stimm abgeben be- 
 schaftigt war." 
 
 '^^ See Lindsay, Syntax of Plautus. 120; Bennett, Syntax of Early 
 Latin, i. 133: Holtze, 2.125. 
 
58 ArcJmism in Aulus GelHus 
 
 prospicienter animadversas in mentes animosque hominum cum 
 audiendi quadam inlecebra induit; 5. 10.12 Sed maius mihi in 
 ista victoria prolubium est, cum te non in causa tantum, sed in 
 argumento quoque isto vinco; 6.3.25 non culpa tantum vacat, 
 sed dignus quoque laude admirationeque est, cum et ingenue 
 ac religiose dicere visus est contra Rodienses, quod sentiebat 
 et . . . ; 6.2.7 ecquale putat cor habere me et quam stultum 
 esse me credit, cum id mihi persuadere vult? 1 1.8.4 Ea cum 
 legisset M. Cato "Ne tu" inquit "Aule, nimium nugator es, 
 cum maluisti culpam deprecari quam culpa vacare" ; 12. 12.4 
 "dKoivovor;Toi" inquit "homines estis cum ignoratis prudentis 
 et cauti patrisfamilias esse, quod emere velit, empturum sese 
 negare propter competitores emptionis. In all these cases, save 
 2.29.1, the tense used by Gellius in the main clause makes it 
 very difficult to explain the c/< ///-clauses as other than causal. 
 
 In Gellius there are several instances of the indicative in de- 
 pendent questions. The indicative in such questions preceded 
 the subjunctive.^*^ In general, the principle of para- 
 taxis^®* accounts for the use of the indicative mood in such 
 questions, especially in conjunction with the imperative. In 
 Plautus the indicative in this type of question is as common as 
 the subjunctive.^*'' Already in Gellius's time the subjunctive 
 was fast losing ground, so that when the Latin language finally 
 broke up into the Romance tongues the use of the subjunctive 
 was greatly restricted. ^^° We see this tendency in Apule- 
 ius :^^^ compare e. g. Ap. Met. 2.22 nee satis quisquam definire 
 poterit quantas latebras . . . comminiscuntur ; Tertullian, Ux. 
 
 "' See e. g. Bennett, Syntax of Early Latin, 1.120. 
 
 "* See e. g. Knapp's Vergil, Index, under Moods 1.4, with notes on 
 the passages there cited; Frobenius, § 186; Draeger, 2.460 ff. ; Gilder- 
 sleeve-Lodge, 467, N. i ; Schmalz, 516. 
 
 "° See Lindsay, Syntax of Plautus, 66. 
 
 ""See Grandgent, Vulgar Latin, § 117 (p. 53): "In conditions not 
 contrary to fact, in indirect discourse and indirect questions, in de- 
 pendent clauses that are not adversative nor dubitative, the indicative 
 was often substituted for the subjunctive" (see the references given by 
 Grandgent) ; Schmalz, 517. Here again the plebeian and the archaic 
 are at one : see above, pages 9-10. 
 
 "' Hoppe, 72. 
 
Archaism in .luliis Gclliiis 59 
 
 1.8 vide, quam ex aequo habetur qui . . . benefecerit; Car. 
 Chr.3 quid tanti fuit edoce.^^- 
 
 Coming now to Gellius himself, let us compare 19.8.6 prop- 
 terea peto ut, si Gai Caesaris liber prae manibus est, promi 
 iubeas, ut quam confidenter hoc indicat aestimari a te possit; 
 2.21.4 Quin . . . vos opici dicitis mihi, quare, quod a/xa^av 
 Graeci vocant, nos septentriones vocaimis ;'^^^ 9.2.5 Quaeso 
 autem te . . . quibus nos uti posse argumentis existimas ;^®* 
 18.7.2 Quaeso te, magister, dicas mihi, mini erravi, quod, cum 
 vellem S77/x77yopt'as Latine dicere, contiones dixi?Here, appar- 
 ently, Gellius regards num erravi as independent, but in 20.1.8 
 we have Die enim, quaeso, die. . .an. . .existumes. In 19.8.6 
 quae ratio est quamobrem C. Caesar vel dictam esse a veteribus 
 vel dicendam a nobis non putat, it may be said that quamobrem 
 is treated as a relative. Since the indicative in indirect questions 
 is common in late Latin, it would not be safe to attribute its 
 use by Gellius to intentional archaism. We have seen that it 
 is frequent in Plautus and the early writers in general. We 
 have also seen that it grows more and more frequent from 
 Gellius's time to the beginnings of the Romance languages. ^^' 
 In any case, however, we have either a reversion to type or a 
 persistence of type, so that the matter properly finds place in 
 this paper. 
 
 A matter more or less akin to that just discussed, though 
 not involving the use of moods, may be noted here, utrum 
 . . . ne . . . an, found in 2.2.7 ^"<^' 7-^-3' is clearly an archa- 
 ism.^*' The passages run as follows: "Absque praeiudicio" 
 
 '" It may be noted that the indicative in these two instances from 
 Tertullian occurs in connection with the imperative ; it is precisely in 
 connection with the imperative that the question could most easily 
 remain more or less independent (paratactic) ; precisely in connec- 
 tion with the imperative, too, occur most of the examples of the 
 (apparent) dependent question in the indicative in early Latin. See 
 Knapp, A. J. P., 32.33, in a review of Bennett's Syntax of Early Latin, 
 Volume I ; Gildersleeve-Lodge, 467, X. 3. 
 
 '" Since the main clauses are logically imperative in force, see above, 
 note 192. 
 
 "' See above, note 192. 
 
 '"On the subject in general see Schmalz, 516-517. 
 
 '"* See Lindsay, Syntax of Plautus, 119; Bennett, Syntax of Early 
 
6o Archaism in Aulus Gellius 
 
 inquit Taurus, "tu interea sede, dum inspicimus quaerimusque,, 
 utrum conveniat tene potius sedere, qui pater es, an filiuni, qui 
 magistratus est; Lepide igitur agitari potest, utrum videri 
 continentiorem par sit Publiumne Africanum Superiorem 
 ... an regem Alexandrum . . ." In the former passage we 
 have another archaism in absque: see Drisler Studies, 167. 
 Cf. Plautus Most. 681 ; Stich. 703; Bacch. 500-501, Trin. 306; 
 Pseud. 709; Mil. 345; Rud. 104; Terence Eun. 721 ; Adel. 382. 
 So far as I have noticed, Professor Bennett does not, in his 
 Syntax of Early Latin: Volume i, discuss the occur- 
 rences of the indicative in subordinate clauses in the oratio 
 obliqua in early Latin.^^^ As will be seen from the references 
 given in note 197 below, the usage does occur in early Latin, and 
 is characteristic of late and plebeian Latin. In the examples 
 cited below, one would expect the subjunctive. But even in 
 the Latin of the best period instances of the indicative occur^^^ 
 which are almost as difficult to account for. Whether the sub- 
 junctive or the indicative shall be used seems to depend upon 
 some subtle feeling which the writer may have at the moment, 
 which in his own mind seems to separate the clause from the 
 oratio obliqua, by way of parenthetical remark, or for special 
 emphasis. As in the case of dependent questions, the occur- 
 rences of the indicative in Gellius may be due either to the 
 (less exact) usage of his times or to intentional archaism. ^^^ 
 
 Latin, 1.333 (§§ 9, u; Bennett's citations are not exhaustive) ; Madvig, 
 Latin Grammar, § 452.1; Brix-Niemeyer', Hallidie, and Lindsay on 
 Capt. 268; Wagner on Aul. 427; Lorenz on Pseud. 688. The matter 
 discussed above must not be confused with quite different expressions 
 in Cicero, in which after a clause beginning with utrum (the neuter 
 pronoun) there is a pause, and a fresh start is made with -;;e . ... an 
 clauses: see Reid on Cic. Acad. 2.71. 
 
 '"For this use see Holtze, 2. 116-117 (§ 4d) : Draeger, 2.444.; Lind- 
 say, Syntax of Plautus, pages 65 f . ; Elmer on Terence Phormio, 9,17; 
 Grandgent, § 117; Gorges, 43-45. 
 
 "' Draeger, 2.442-443. 
 
 '^' Apuleius has similar constructions : for example, see Met. 2.21 
 ignoras Thessaliae te consistere, ubi sagae mulieres .... demorsitant; 
 Met. 3. 6 non tantum impunem me, verum etiam laudabilem publice 
 credebam fore qui ... . apud meos semper innocentiam commodis 
 cunctis antetuleram. 
 
Archaism in Aulns GcJlius 6l 
 
 A few examples from Gellius will suffice by way of illustration. 
 Compare 1.12.10-11 De more autem rituque capiundae virginis 
 litterae quidem antiquiores non extant, nisi quae capta prima 
 est a Numa rege esse captam. Sed Papiam legem invenimus, 
 qua cavetur, ut pontificis maximi arbitratu virgines e populo 
 viginti legantur sortitioque in contione ex eo numero fiat et, 
 cuius virginis ducta crit, ut eam pontifex maximus capiat . . .; 
 5.18.6 historias quidem esse aiunt rerum gestarum vel exposi- 
 tionem . . . annales vero esse, cum res gestae . . . componuntur; 
 6.1.2-4 Nam et C. Oppius et lulius Hyginus aliique. . .tradunt 
 . . . neque multis diebus, postquam ille anguis in lecto visus 
 est, mulierem coepisse . . . ; 6.3.19 Quippe recte et utiliter . . . 
 praecipitur indices de capite alieno deque causa ad sese non 
 pertinenti cognituros, ex qua praeter officium . . . nihil . . . 
 redundaturum est, conciliandos esse . . . existimationi saluti- 
 que eius, qui apud eos accusatus est (possibly the indicative 
 here and in the following example is an echo from Cato : the 
 whole chapter deals with a speech of Cato) ; 6.3.22 Sed quod 
 ait confessum Catonem noluisse Rodiensis ita depugnari ut 
 depugnatum est; 5.1 i.i Existimant quidam etiam illud Biantis^ 
 . . . responsum consimile esse atque est Protagorion illud ; 
 6.17.3 Quis adeo tarn linguae Latinae ignarus est, quin sciat 
 eum dici obnoxium, cui quid ab eo, cui esse obnoxius dicitur, 
 incommodari . . . ; 10.2.2 Sed et divo Augusto imperante, qui 
 temporum eius historiam scripserunt, ancillam ... in agro 
 Laurente peperisse . . . dicunt . . . matrem . . . non multo, 
 postquam peperit, mortuam . . . ; 14.1.2 eaque fuerunt ad 
 banc ferme sententiam : disciplinam istam Chaldaeorum tantae 
 vetustatis non esse, quantae videri i-oluiif, neque eos principes 
 eius auctoresque esse, quos ipsi ferant (note the shift in 
 moods, voliint . . . ferant) ; 14. 1.20 declarari . . . oportere 
 dicebat . . . quales . . . homines gignerentur . . . quinam 
 olim futuri essent . . . ut . . . stellae istae praemonstrare 
 debuerint, qualis qualique fato futurus sit, quisquis hodie 
 flatus est. 
 
 Gellius is fond of repeating atque, a trick of style apparently 
 borrowed from Cato.-"° For examples see 1.23. i cum nuilta 
 
 "* See Draeger, 2.54; Schmalz, 496 (§ 242); Minton Warren. P. A. 
 
y/ 
 
 62 Archaism in Aulits Gcllius 
 
 quidem venustate atque luce atque munditia verborum ; 2.2.9 ^^ 
 publicis locis atque muneribus atque actionibus ; 2.8.1 inperfecte 
 atque praepostere atque inscite ; 6.3.52 distincte nimis atque 
 compte atque modulate; 6.1 9.1 Fulcrum atque liberale atque 
 magnanimum factum; 10.3. 13 impense atque acriter atque in- 
 flammanter; 12. 1.9 filium proprii atque consueti atque cogniti 
 sanguinis alimonia privare; 17.1.1 parum integre atque in- 
 proprie atque inconsiderate locutum.^°^ 
 
 sed enim occurs e. g. in Praef. 18; 1.7.18 sed enim praeter 
 Plautum . . . multam . . . copiam offendimus . . . ; 2.6.10; 17. 1.3 
 (see Apul. Met. 1.24). This may be an archaism. Draeger 
 333.2c, says : "Zuerst Cato Orat. p. Rhod. Dann Cic. p. Gael. 
 24,60 . . . Mehr wird aus Cicero nicht angefiihrt, nichts aus 
 Caesar, Sallust und Livius ; mehr schon aus den Augusteischen 
 Dichtern . . . Im silbernen Zeitalter scheint es zu fehlen, sicher 
 bei Tacitus, dann kommt es wieder zum Vorschein". P. Langen 
 (Beitrage zur Kritik und Erklarung des Plautus, page 263) : 
 "ob Plautus auch sed enim gesagt habe, ist zweifelhaft, iiber- 
 liefert ist kein Beispiel der Art". Lindsay (Syntax of Plautus, 
 page 97) writes: "but sed enim seems not to occur (in Bacch. 
 1080 the MSS offer et enim ...)", Leo, it should be noted, 
 reads there sed enim, Goetz-Schoell at enim. Schmalz, 508, 
 declares that sed enim "schon vor Cato . . . gerne . . . von den 
 aug. Dichtern und den Archaisten, z. B. Pronto, Gellius ge- 
 braucht wird". 
 
 Twice after a temporal clause Gellius introduces what is 
 logically the main clause with a copulative conjunction {atque, 
 et).^°- According to Schmalz this illogical form of parataxis is 
 
 P. A., 25. xliv (1894). Elmer, in his dissertation on Que, et, atque in 
 the Inscriptions of the Republic, in Terence and in Cato (see A. J. P., 
 8), did not notice Cato's fondness for repeated atque. 
 
 *°^ Compare Pronto 36 (Naber) : Uni M. Porcio me dedicavi atque 
 despondi atque delegavi. Hoc etiam ipsum atque unde putas? ex ipso 
 furore. Lodge, Lexicon Plautinum, s. v. Atque, 17, gives examples of 
 varying combinations into which atque enters twice, such as atque . . . 
 atque; atque . . . atque . . . -que; atque . . . -que et . . . atque; et . . . atque 
 .... atque. 
 
 °" For this use in Plautus see Lodge, Lexicon Plautinum, s. v. Atque, 
 16 (page 179). 
 
Archaism in Aldus Gellius 63 
 
 confined to two authors only, Plautus and Gellius; see 
 497 (§ 244). Compare 2.29.8 Haec tihi ille dixit et discessit; 
 17.20.4 Haec verba nbi lecta sunt atque ibi Taurus mihi . . . 
 inquit. It should be noted that 2.29 is a perfect mine of ar- 
 chaisms, such as cum causal with indicative, fervit, crastini, luci, 
 etc., etc.^"' The Plautine examples of the special phenomenon 
 under discussion are Epid. 217 (Schmalz wrongly cites as 209) 
 Quom ad portam venio atque ego illam illi video praestolarier ; 
 Bacch. 278 forte ut adsedi in stega, dum circumspecto, atque 
 ego lembum conspicor; Most. 1050 Quoniam convocavi, atque 
 illi me ex senatu segregant ; Poen. 649 Nescimus nos quidem 
 istum qui siet; Nisi dudum mane ut ad portum processimus, 
 Atque istum e navi exeuntem oneraria Videmus; Merc. 256 
 Postquam id quod volui transegi, atque ego conspicor Navem 
 ex Rhodo quast heri advectus filius. 
 
 Finally, in 2.29, we have convincing evidence of Gellius's 
 conscious imitation of the ancients, whom he so profoundly 
 reverenced. In his rendering of Aesop's fable, we have numer- 
 ous examples of archaisms of form and of syntax,-"* which, 
 without doubt, are taken from the Latin version which Gellius 
 was following. But unmistakable evidence that Gellius's work 
 is, in part, a loose paraphrase of one of the Satires of Ennius,^"' 
 and, in part, a more or less faithful rendering of parts of verses, 
 or even of whole verses, of his Latin original, is found in the 
 bits of trochaic lines which characterize this fable-°^ as given 
 us by Gellius. 
 
 Trochaic rhythm is natural enough to Latin, but here we 
 have no accidental measures due to the nature of the language. 
 By way of contrast it is suggestive to compare the story of 
 the lark, 2.29, with the story of Arion, 16.19. So far as I have 
 been able to discover, the striking archaic and metrical ele- 
 ments which are so remarkable in the former are lacking in the 
 
 *" See Knapp, Drisler Studies. 143-144 (for archaisms of form), 
 and, for archaisms of synta.x, above, page 57. The next paragraph of 
 this page may also be compared here. 
 
 "^ See above, note 203. 
 
 "Cf. § 20. 
 
 "* See Vahlen's Ennius' (1903), ccxxii-ccxxiii, and Knapp, A. J. P., 
 32.28. 
 
64 Archaism in Aulus Gellius 
 
 latter. Apparently, Gellius is here translating from Herodotus, 
 and so is himself responsible for the Latin. There was no in- 
 termediary archaic Latin version from which to borrow. There 
 seem to be no reminiscences of Ovid's rendering (Fasti, 
 2.79-1 18). Ovid, it may be noted, is never mentioned in the 
 Noctes Atticae (compare above, page 20). 
 
 To the trochaic rhythms which have already been noted by 
 Vahlen-°'^ and by Knapp,-°^ in 2.29, the following, not in all 
 cases, to be sure, very musical, should be added : appetat messis 
 pullis iam iam plumantibus ; flavescentibus pullis ; ipsa iret 
 cibum pullis ; postea segetum illarum ; amici isti magnam par- 
 tem inquit ; temperi ad metendum ; afferes primo luci. 
 
 ^' Ennius, ccxxii. 
 *" A. J. P., 32.30-31. 
 
SELECTED BIBLIOGRAPHICAL INDEX^ 
 
 Abbott. F. F. The Use of Repetition in Latin to Secure Emphasis, 
 Intensity, and Distinctness of Impression, Studies in Classical Phil- 
 ology, 3.67-87 (1900, Chicago). 
 
 A. J. P. American Journal of Philology. 
 
 Archiv. Archiv fiir lateinische Lexicographic und Grammatik. 
 
 Bennett, Charles E. Syntax of Early Latin, Vol. I. The Verb (Bos- 
 ton, 1910). 
 
 Bernhardy, Gottfried von. Romische Litteratur (Halle, 1872). 
 
 Brock, M. Dorothy. Studies in Fronto and His Age (Cambridge, 
 1911). 
 
 Biicheler, Franz. Grundriss der lateinischen Deklination (Bonn, 1879). 
 
 Cooper, Frederic Taber. Word Formation in the Roman Sermo 
 Plebeius (New York, 1895). 
 
 Desertine, A. H. J. V. M. De Apulei Studiis Plautinis (Neomagi, 
 1898). 
 
 Draeger, A. Historische Syntax der lateinischen Sprache. (2 vols. 
 
 Leipzig, 1878-1881). 
 Duff, J. W. A Literary History of Rome (London and Leipzig, 1909). 
 Ellis, Robinson. Commentary on Catullus' (Oxford, 1906). 
 Frobenius, Rudolf. Die Syntax des Ennius (Nordlingen, 1910). 
 Georges, K. E. Lateinisch-Deutsches Hand-Worterbuch' (Leipzig, 
 
 1899)- 
 Gildersleeve, Basil, and Lodge, Gonzales. Latin Grammar (New York; 
 
 1900). 
 Gorges, O. De Quibusdam Sermonis Gelliani Proprietatibus Observa- 
 
 tiones (Halle, 1882). 
 Grandgent, C. H. An Introduction to Vulgar Latin (Boston, 1907)- 
 Gronovius, J. Aulus Gellius (Lyons. 1706). 
 Haupt, Moritz. Opuscula (Leipzig. 1875-1876). 
 Helm, Rudolphus. Apulei Opera quae supersunt (Leipzig, 1907). 
 Hertz, M. (Vind. Gell. Alt.). Vindiciae Gellianae Alterae (Leipzig, 
 
 1873). 
 
 A. Gellii Noctes Atticae (2 vols. Berhn. 1885). 
 
 Opuscula Gelliana (Berlin, 1886). 
 Hildebrandt, G. F. Apulei Metamorphoses (Leipzig, 1862). 
 Holtze, F. Guilelmus. Syntaxis Priscorum Scriptorum Latinorum 
 
 (Leipzig, 1861). 
 Hoppe, H. Syntax und Stil des Tertullian (Leipzig, 1903). 
 Hosius, Carolus. A. Gellii Noctes Atticae (Leipzig, 1903). 
 Knapp. Charles. Archaism in Aulus Gellius (in Classical Studies in 
 
 Honour of Henry Drisler, New York, 1894)- 
 
 Corrections and Additions to Lewis and Short in Connection with 
 
 Aulus Gellius, A. J. P. 14- 216-225. 
 1 See note .3. 
 
66 Selected Bibliographical Index 
 
 Notes on the Prepositions in Gellius, T. A. P. A. 25 (1894). 1-33. 
 
 Stories from Aulus Gellius (New York, 1895). 
 
 The Originahty of Latin Literature, The Classical Journal, 3. 251-260, 
 
 299-307. 
 
 Vahlen's Ennius, A. J. P. 32. 1-35. 
 Kretschmann, Henricus. De Latinitate L. Apulei Madaurensis 
 
 (Konigsberg, 1865). 
 Kretschmer, J. De A. Gellii Fontibus (Posen, i860). 
 Kiibler, Bernhard. Die lateinische Sprache auf africane Inschriften 
 
 (in Archiv 8 [1893]). 
 Kiihner, Raphael. Ausfiihrliche Grammatik der lateinischen Sprache 
 
 (Hanover, 1 877- 1879 ) ; second edition (2 vols. Hanover, 1912). 
 Landgraf. See Stolz, Landgraf. 
 Langen, P. Beitrage zur Kritik und Erklarung des Plautus (Leipzig, 
 
 1880). 
 Leky, Maximilian. De Syntaxi Apuleiana (Bonn, 1908). 
 Lindsay, W. M. The Latin Language (Oxford, 1894). 
 
 Syntax of Plautus (Oxford, 1907). 
 Lodge, Gonzalez. Lexicon Plautinum (Leipzig, 1901-1911). 
 Lodge, Henry Cabot. Shakespeare's Americanisms (Harpers Monthly, 
 
 January, 1885). 
 Mackail, J. W. Latin Literature (New York, 1907). 
 Madvig, J. N. Latin Grammar, revised by Thomas A. Thacher (Bos- 
 ton, 1888). 
 Mohl, F. G. Introduction a la Chronologic du Latin Vulgaire (Paris), 
 
 1899). 
 Monceaux, Paul. Histoire litteraire de I'Afrique chretienne (Paris, 
 
 1901-190S). 
 Naber, S. A. M. Cornelii Frontonis et M. Aurelii Imperatoris Episto- 
 
 lae (Leipzig, 1867). 
 Nagelsbach, K. F. Lateinische Stilistik fiir Deutsche' (Niirnberg, 
 
 1888). 
 Naudet, Joseph. M. Accii Plauti Comoediae, Lemaire edition (Paris, 
 
 1830). See Bibliotheca Classica Latina, Vols. 68-71. 
 Nettleship, Henry. Lectures and Essays (First Series, Oxford, 1889). 
 
 Lectures and Essays (Second Series, Oxford, 1895). 
 
 Contributions to Latin Lexicography (Oxford, 1889). 
 Neue-Wagener = Neue, C. F., and Wagener, C. von. Formenlehre der 
 
 lateinischen Sprache' (2 vols. Berlin, 1892-1902). 
 Norden, E. Die antike Kunstprosa (Leipzig, 1898). 
 Olcott, George. Studies in the Word Formation of the Latin Inscrip- 
 tions (Rome, 1898). 
 Piechotta, James. Curae Apuleiamae (Breslau, 1872). 
 Priebe, C. De Frontone imitationem prisci sermonis latini adfectante 
 
 (Settin, 1885). 
 
Selected Bibliographical Index 67 
 
 Purser, Louis C. The Story of Cupid and Psyche as Related by Apu- 
 
 leius (London, 1910). 
 Reisig-Haase = Reisig, Christian Karl, and Haase, Frederich. Vorlesun- 
 
 gen iiber latcinische Sprachwissenschaft. Revised by Hagen, Heer- 
 
 degen. Schmalz and Landgraf (2 vols. Berlin, 1881-1890). 
 Ribbeck, Otto. Tragicorum Romanorum Fragmenta' (Leipzig, 1897- 
 
 1898). 
 Riese, F. A. Varro, M. Terentius, Saturarum Menippearum Reliquiae 
 
 (Leipzig, 1865). 
 Roby, Henry J. Grammar of the Latin Language from Plautus to 
 
 Suetonius (London^ Volume i, 1887, Volume 2, 1874). 
 Schmalz. See Stolz, Schmalz. 
 
 Sellar, W. Y. The Roman Poets of the Augustan Age (Oxford, 1892). 
 Steele, R. B. Archaisms Noted by Servius in the Commentary on 
 
 Vergil, American Journal of Philology, 15 (1894). 164-193. 
 Stolz. See Stolz, Schmalz. 
 Stolz, Landgraf = Stolz, Friedrich, and Landgraf, Gustav. Historische 
 
 Grammatik der lateinischen Sprache (vols, i and 3. Leipzig, 1894, 
 
 1903). 
 
 Stolz, Friedrich, Schmalz, J. H. Lateinische Grammatik und Stilistik*, 
 in Iwan v. Miiller's Handbuch der klassischen Alterumswissen- 
 schaft, Volume 2 (Munich, 1910). 
 
 Teuffel-Schwabe-Warr. A History of Roman Literature, 2 vols. (Lon- 
 don, 1900). 
 
 Thesaurus. Thesaurus Linguae Latinae (Leipzig, 1900-). 
 
 Thomas, P. Apuleius, De Philosophia Libri (Leipzig, 1908). 
 
 T. A. P. A. Transactions of the American Philological Association. 
 
 Vahlen, Johannes. Ennianae Poesis Reliquiae* (Leipzig, 1903). 
 
 Vogel, Theodor. De A. Gellii vita, studiis, scriptis narratio et iudicium 
 (Zittau, i860), 
 
 Weiss, Fritz. Die Attischen Nachte des Aulus Gellius (2 vols. Leipzig. 
 
 1875). 
 Wolfflin, Eduard. Lateinische und romanische Comparation (Erlan- 
 gen, 1879). 
 Bemerkungen iiber des Vulgarlatein. Philologus, 34.149. 
 
VITA. 
 
 I, Walter Eugene Foster, was bom in Westmoreland, New- 
 Hampshire, May 13, 1870. I prepared for college at Gushing 
 Academy, Ashburnham, Mass., from which I was graduated 
 in 189 1. After four years of teaching, I entered Williams 
 College, from which I received the degree of B.A. in 1899. 
 My graduate work has been done at Columbia and Cornell 
 Universities. 
 
 I have held the following educational positions : principal 
 of the Pratt High School, Essex, Conn. ; teacher in, and later 
 principal of, the Williamstown High School, Williamstown, 
 Mass.; teacher of Latin in the Morris High School, New 
 York City. At the present time I am head of the department 
 of Latin in the Stuyvesant High School, New York City. 
 
 To Professors James C. Egbert and Nelson G. McCrea 
 I wish to express my thanks for their unfailing kindness and 
 consideration. Especially do I wish to thank Professor Charles 
 Knapp, to whose suggestion this dissertation is due, and to 
 whose kindly interest and wide scholarship I have never ap- 
 pealed in vain. 
 
RETURN CIRCULATION DEPARTMENT 
 
 T O r 202 Ma i n Library 
 LOAN PERIOD I 1^ 
 
 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 
 
 DEC0419BA 
 
 p^fFAQjTAMPED BEL OW . 
 
 UNIVERSITY OF CALIFORNIA, BERKELEY 
 FORM NO. DD6, 60m, 1 /83 BERKELEY, CA 94720 
 
UC. BERKELEY LIBRARIE 
 
 C0DM0T7343