E48 Cornell University Library PA 2277.E48 The copulative conunctons que et, ati 3 1924 021 616 689 The Copulative Conjunctions ^ue, Et, Atque s in the Inscriptions of the Republic^ in Terence^ and in Cato. r By H^ C: ELMER, Ph. D., Late Fellow and Assistant in Latin in the Johns Hopkins University, A Disseriaiion presented for the degree of Doctor of Philosophy. Reprinted from the American Journal of Philology, VoL VIII, No. 3. BALTIMORE : Press of Isaac Fhiedenwald, 1887. £1^7 TABLE OF CONTENTS. PAGE. Introductory, 3 Ways of combining sentences 4 Alque (ac') rare in inscriptions, 4 " " before vowels and consonants, 4 Attempt to make que the universal connective in the laws, . 5-8 Et more common in other classes of inscriptions; inaccuracy of dictionaries on this point, ■ 8-9 Signification of que, et ; criticism of definitions as given by Ballas, Drager, etc 9-10 ^«« with words in final ) contrasted ; {c) special added to general ; (rf) adding reason ; (e) adding accessory notion, . . . . . . . .34 24. Hendiadys, .......... 35 25. Polysyndeton, 35-37 26. Adding last term, ........ 37 27. Correlative uses, 38 28. Sporadic uses, 38-39 QUE, ET, ATQUE IN THE INSCRIPTIONS OF THE REPUBLIC, IN TERENCE, AND IN CATO. I. Introductory. In preparing the following paper I have made the inscriptions the central point of investigation, using Terence and Cato chiefly by way of comparison and contrast, as convenient representatives of poetry on the one hand and prose on the other. Since the publication of the Corpus Inscriptionum Latinarum, Vol. I, other inscriptions dating from republican times have been discovered and edited in Ritschl's Supplements, in the Hermes, and in the Ephemeris Epigraphica. It was my original intention to embody the conjunctions que, et, atque from all of these in this paper, but I have as yet been able to collect only those from the first Volume of the Corpus and from the Ephemeris Epigraphica, the last mentioned presenting only the one instance of which. I have spoken in the third division of this paper. The additions which might be made from the other sources are very few, and they could not make any material difference with the results. In Cato I have found it necessary to refer to pages and lines, instead of chapters and sections, as a section often includes many instances of the conjunctions, and there would be no convenient way of designating which particular one was referred to in each case. In doing this I have used Keil's edition of de Agri cultura (Leipzig, 1882), and Jordan's collection of the fragments of Cato. (Leipzig, 1862). For Terence I have used the standard edition, of Umpfenbach. This paper is intended to be used side by side with the literature that has hitherto appeared on the copulative conjunctions, and I have, accordingly, introduced but little from other papers into my own. I have quoted them, however, not infrequently, for the pur- pose of modifying or correcting some of their statements. Where I have adopted the views or statements of others, credit is duly given in the foot-notes or in parentheses. Throughout the paper C. represents the first volume of the Corpus Inscriptionum Latinarum. 4 Que, Et, Atque in the Inscriptions Sentences may be put together in connected discourse in two ways, paratactically or hypotactically. They may be ranged side by side, each occupying a position equally prominent with the others, leaving their connection and relative importance to be inferred from the nature of the thoughts ; or they may be reduced to syntactical unity by grammatical subordination. Again, in the former case they may be left wholly without a connective, or they may be joined together by what have been called the " paratactic conjunctions." The most common form of such parataxis is that of copulation, and it is with the particles which find application here that the present paper has to do. These are et, que, atque (jic). Of these three conjunctions que is of the most ancient formation, and is the only one found in those inscriptions which are prior to the Second Punic War (comprising the Pars Prior of the first volume of the Corpus), with the exception of " Erinie et . . ." in a fragment (C. 182) and atque in one of the Scipio inscrip- tions (C. 33). In the inscriptions under consideration atque (ac) is compara- tively rare,' though it occurs on earliest and latest monuments alike, and seems to be quite equally distributed over all periods and classes of inscriptions, appearing, for instance, in the epitaph of P. Cornelius Scipio (C. 33), then three times in the S. C. de Baccha- nalibus (C. 196, B. C. 186), once in the Lex Agraria (C. 200, B. C. Ill), three times in the Lex Rubria (C. 205, B. C. 49), and five times in the tituli. In all but three instances ' the word following atque begins with a vowel. In the three exceptions it begins with m, p, s, respect- ively. The contracted form ac is not found earlier than the Lex Antonia (C. 204, B. C. 71), and occurs only seven times in all, once before each of the letters d, f, m, n, p, and twice before s.' The examples of atque (^ac') in the inscriptions are not numerous enough to afford sound basis for comparison with the use else- where. But it is interesting to note, in this connection, that Cato uses ac only three times,' once before each of the letters m, v, d, while he uses atqice (gi times in all) indiscriminately before any letter (vowel or consonant) except u. '^ Atque 13 times, 9 times with copulative force; ac 7 times, 6 times with copulative force. - C. ig6, ig, atque mulieres ; 205, I 18, atque sei ; 1480, atque propinqueis. * C. 204, I 20, ac ne ; 205, II 47, ac sei ; 1007, i, ac pellige ; 1008, 3, ac finem • loo8, 9, ac s. ; ioo3, 14, ac v. ; iot2, 7, ac m. ■•De Agr. cult., ch. 41, i, ac maloiiim ; Fragmenta (Jordan), p. 36, 8 ac vicissim; 66, 9, ac ducentis. of the Republic, in Terence, and in Caio. 5 In contrast with Cato, Terence has a distinct use for each form of this conjunction, viz. atque before vowels and before h, and ac before consonants. Of the 210 instances of atque in Terence there are only 7 exceptions to this rule, atque being used once before each of the letters g and c, twice before 1, and three times before consonantal u. Of ac there are 66 instances in Terence, occurring, as elsewhere, only before consonants. It is in most cases employed even before the letters just given where atque sometimes occurs. Drager says : Alle iibrige Beobachtungen sind werthlos, ausgenommenon etwa die, dass Plautus ac nicht vor q stellt. He might have added that Terence also does not allow ac before q. The most common of the copulative conjunctions are que and et, and between these two the inscriptions present striking differ- ences. In the legal enactments here preserved, et is avoided and qtie is used indiscriminately in all connections. This is clearly seen in the Lex Agraria (C. 200), where que appears 46 times and et only once. Again, in the Lex Agraria Repetundarum (C. 198), que is used 39 times and et only 9. In the Lex Cornelia (202) et does not occur at all, while que appears 25 times. In the laws, decrees and legal forms recorded in the inscriptions and com- prising in the Corpus Nos. 196 to 198, 200, 202 to 211, there are, in all, 210 instances of que and only 43 of et. Following is a table showing the number of times that et and que occur in each of these : ■ C. 196 De Bacchanalibus que ^ et o 197 Lex Reperta Bantiae .... 4 2 198 Lex Acilia Repetundarum . . 39 9 200 Lex Agraria 46 i 202 Lex Cornelia 25 o 203 S. C. de Asclepiade .... 5 2 204 Lex Antonia 22 3 205 Lex Rubria 23 7 206 Lex Julia Municip 37 18 207 Fragment i i 208 " 2 o 209 " I o This remarkable preponderance of que over et led me to suspect that there was an attempt to make que the universal copulative conjunction in the laws, and that, where et, atque, or ac occurs, the 6 Que, Et, Atque in the Inscriptions use of que could not, for some reason, be allowed. With this possibility in view I have examined all the instances of et, atque^ ac in the legal enactments recorded in the inscriptions and have found much to favor the hypothesis. In the Lex Acilia Repetundarum et occurs 9 times. In three instances the phrase is et unde (lines 26, 27, 28). Here the use of que would necessitate appending it to short e, which I have else- where shown is not allowed.' It would also afford the possibility for confusion with the adverb undique. In another case (32) we have et, quom, where the clause introduced by quom is thrown in parenthetically and the use of que would improperly incorporate it into the context. Again (34), et si. There was, for some reason, an aversion to appending que to si.' It is never allowed in the inscriptions, et and ac always being substituted.' Again, in 53, et in. There was a similar aversion to appending ^«if to in, as I have further on in this paper pointed out. This may have been because inque would have coincided in form with the verb. Que could not have been appended to the object of the preposition, as that was of very rare occurrence in early times, and it was not allowed at all in the inscriptions, or in Plautus (Ballas, Gram. Plant., Spec. I, p. 19). Thus, in six of the nine instances there is evident reason for avoiding que. In the Lex Agraria there is only one et,'' and that is another case where que would have made inque. In one of the two cases in the S. C. de Asclepiade we have et ad. Adque might have been mistaken for the conjunction. There was, too, at all periods an aversion to appending que to this prepo- sition (Schmalz, Lat. Synt. §170). , In the Lex Antonia there are three instances of et, but two of them connect proper names opposed to each other (lines 19, 21), inter Romanes et Termenses. Here separation and opposition are to be understood and que would be manifesdy inappropriate. In the other case (29) we have et, quo, to distinguish from qudque; quoque. In the Lex Rubria there are seven instances of et. Three of them (II, 50, 27) are followed by set, and are used, as above pointed out, to avoid seique. Two others (7, 8) are followed by ' See third division of this paper, " Words in final /." ' Possibly because, in the popular pronunciation, siqjie might in some cases have become sic ; cf. neqiie^ ncc, 2 C. 205, I I r, et set ; I 50, et nisei ; II 27, et sei ; II 47, ac sei ; 628, et sei. * C. 200, 30. of the Republic, in Terence, and hi Cato. 7 the preposition a, ab. Drager (Hist. Synt. II, p. 37) has already- pointed out that the language, at all periods, had an aversion for appending que to this preposition. In (42) et municipium. The language at no period allowed, with any freedom, the appending of que to words of five or more syllables.' The laws under con- sideration offer but one such instance." In another case (45) we have et in, again to avoid inque. This accounts for all of the 7 instances. The Lex Julia Municipalis offers the large number of 18 instances of et. But 4 of them are with relative pronouns which, \{ que had been used, would have given rise to such ambiguous forms as quaque, quoque, etc.^ Then, from analogy with the singular, it is not surprising to find quorumque avoided in line 35 {et quorurrC) and quarumque in line 60 (et quaruin). Another is et quot (146), to avoid quotque (similar to quodque). Another (39) would have formed inque; another (67) aque ab{_s)que. Another (15) is fol- lowed by a parenthetical clause, which que would have bound too closely to the context. Again, in (29) inter aedem sacram et aedificium locumve publicum, where the two substantives are opposed to each other. Another similar use of et occurs in the same line, but with repetition of the preposition : inter aedem sacram et inter aedificium privatum. Interque is, moreover, a rare combination, not occurring in the inscriptions or in Caesar, who uses only atque with inter (Ringe, Zum Sprachgeb. des Caesar, p. 19). In 157 et is used adversatively (or where we should more naturally use an adversative conjunction). This accounts for 14 out of the 18 instances. In S. C. de Bacchanalibus, the Lex Cornelia, and in the frag- ments numbered 208, 209 in the Corpus, et does not occur at all. Atque and ac are used only 4 times in the inscriptions recording laws, with copulative force. In two of them (196, 26 and 28) atque utei may have been used to avoid confusion with the form of the adverb utique,* and in 204, 1 20, ac ne locenter, to avoid the ambiguity arising from neque. To sum up, then, of the 47 instances of et, atque {ac) in these legal documents, 33 of them may very probably have been used to avoid the confusion which que would have occasioned with other words, and to avoid unusual combinations. The only instances which remain to be accounted for are : (i) et is, eam, eum, ' Drager, Hist. Synt. II, p. 37, §3. ^ C. 200, 93, possessionesque. 2 5, et quae ; 14, et quo ; 36, et quo ; 35, et quo.. ^Uii-\-que occurs in the same inscription. Despite occasional exceptions there was undoubtedly a general tendency to avoid such ambiguous forms. 8 Que, Et, Aique in the Inscriptions eos, in C. 197, 21 ; 198, 53 ; 198, 76 ; 206, 13 ; 206, 157 ; 207 ; (2) et causam, C. 198, 32; (3) et quantum, C. 198, 48 ; (4) et eidem, C. 197, 19 ; (5) et de, C. 203, 5 ; (6) et praetor, C. 206, 10 ; (7) aed. et IIII vir, C. 206, 50; (8) et rationem, C. 206, 147. A closer study of the language would probably reveal other similar reasons for the use of et and reduce still more these 14 remaining cases where et seems voluntarily chosen as connective. So much for those inscriptions which record legal enactments. As soon, however, as we examine the others, there appear impor- tant differences in the use of these conjunctions. The tituli, the glandes, the tesserae consulares, in fact all the inscriptions pos- terior to the Second Punic War, except those recording laws, decrees, etc., show, on the whole, a preference for et. These comprise the second, third and fourth sections of the Pars Posterior of the Corpus, or from No. 212 to 1499 inclusive (exc. 627, 628), and show a total of que 108 and et 158. Some of these, however, are in legal form, or in connection with legal forms, and in such cases there is a distinct return to que as the connective. For instance, No. 577 of the Corpus (Lex pariete faciendo) has 16 instances of que and only 5 of et, one of these (2, 18) being followed by a word ending in e, and another by a word of six syllables (38, et duoviralium). No. 603 (Leges aedis Jovis, etc.) has 9 instances of que and none at all of et. In general, then, que is preferred only in legal forms, while else- where et is the more common connective. The statement in Harpers' Diet, credited to Drager, that que is preferred in archaic language to et is, as far as the inscriptions are concerned, inexact, and should be modified so as to apply only to those inscriptions prior to the Second Punic War and to legal language. This preference for que as connective in legal documents remained in classical times. Cicero, for instance ' (Phil. 14, 36), in framing a Senatus Consultum, uses que 23 times, et only 3 times (instead of twice, as Drager says), and atque only once (Drager says not at all). In two of the instances of et the following word ends in e : existimare et iudicare,'' salute et liberate; and atque is followed by ita {itaque would be ambiguous) ; all of which goes to strengthen the hypothesis, above laid down, that et was used only where que would have been impossible. Elsewhere Cicero shows ' Drager, Hist. Synt. II, §314, p. 34. ' In the inscriptions, infinitives are always joined by qw. In Cicero's time the final e had become fixed as short, and so, as we should naturally expect que is avoided and et used in its stead. of the Republic, in Terence, and in Cato. 9 a decided preference for et (e. g. Pro. Rose. Am., et 269, que 121, atque 81, ac 47). That the same preference for que in legal language holds good for imperial times is clearly shown in the Lex regia de Vespiani imperio, in which que is continually used in all connections, while et and atque are not used at all, and ac only once.' II. Significations of Que, Et, Atque. Hand, in his Tursellinus, opens his treatise on et with this sentence : Notiones et sententiae aut coniunguntur simpliciter, aut adiunguntur accessione aut iniunguntur aequatione, applying these words to et, que, and a/^?^e respectively. It is, however, impossible, as Spitta remarks," to formulate a rule which will universally hold. Each author has peculiarities of his own in his use of the copula- tive conjunctions. Plautus, for instance, while in general assigning each particle its peculiar province, still uses them very often indis- criminately.' Seneca, who depended for effects not so much upon a well-rounded whole as upon the individual charm of his short incisive sentences, prefers el except in comparatively rare cases. Klammer' finds that, in his letters, Seneca uses et 3500 times, while que occurs only 550 times, ac 350, atque 100. Curtius, on the other hand, seems to have used et and que with about equal frequency.' Legal language offers the other extreme. A law is looked upon as one closely connected whole and its provisions are accordingly linked together with que. There is nothing in the nature of et and que to force the use of one invariably in a certain connection and the other in a certain other connection. Grammars tell us that et connects in the most general way, without any additional signification whatever, while que implies that the things belong closely to one another and that the second member completes or extends the first. But there is no reason why the things thus connected by que cannot be conceived of each by itself and connected by et. Again, if two things are closely related to each other, and are set side by side, the use of ^/ cannot interfere with the relation between them ; they will be quite as intimately associated as when connected by que. When, in C. 547, exemplum a has Atestinos et Patavinos, while exemplum b has Atestinos Patavinosque, we are not conscious of any real difference between the two, and must regard the 1 Drager, Hist. Synt. II, §314, p. 34. ^ De Taciti in componendis enuntiatis ratione, pars prior, p. 39. ' Ballas, Gram. Plaut. Spec. I, de particulis copulativis, p. 3. ■•Animadversiones Annaeanae Grammaticae, Bonn, 1878, p. 2. lo Que, Ei, Atque in the Inscriptions choice of connective simply as a matter of taste. With this may be compared the example given by Schmalz ' from Tac. Ann. I, i, Tiberii Gaiique et Claudii ac Neronis, where the connectives seem chosen quite arbitrarily. But in enumerating the parts of a whole, where each is distinct in itself, the associations of que would make this particle manifestly inappropriate. So when Varro (Ling. Lat. 8, i) says : quod est homo ex corpore et anima, he was allowed no choice of connectives. Many have made the mistake of assigning to each of these particles various meanings. For instance, Ballas (De particulis copulativis, p. i6, 6), commenting on Plant. Amph. 15 : vicit Per- duellis et domum laudis compos revenit, says that et here has the force of et ideo. He has merely compared the parts connected and made et responsible for the result. So Drager (Hist. Synt. II, p. 65, §317) says: et so wohl als atque dient zur Betonung des Folgenden. This seems to me wholly wrong. The part added may be emphatic, but et has nothing to do with making it so. It would be quite as logical to say that, in the English phrase " go and see," "and" means " in order that," because the latter verb expresses the purpose of the former. Take, for instance, Ter. Adelphoe 648, opinor et certo scio, where Drager and others would assign an emphatic meaning to et. Such a meaning is assigned to it only because, in this particular case, the words following et are more emphatic than those preceding. If they happened to be less emphatic, one might equally well claim for et the opposite meaning. Et, in fact, when used to connect words and clauses, never means more than simple and. It may be preferred where the parts connected have certain relations to each other, and these relations may often be more clearly brought out in translation by an additional word, but the relation is not expressed by the con- junction itself.^ The same is true of que ; though implying a closer union, it by no means expresses any of the varied relations which the added notion may have to the preceding. III. Que AND Words in Final k. It is well known that que, in classical Latin, is not appended to words ending in short e. To this there is, according to Harant 1 Lat. Synt. §175. Poets and later prose writers allow a free interchange of et, atque, que; cf. Reisig, Lat. Sprachwissenschaft, p. 195, §232 (406c). ^ Certain phrases may become fixed, and custom may favor the use of but one connective in such phrases, e. g. senatus populusque Romanus. But here again the occasional use of ^/ shows that no real difference in meaning is felt between them ; cf. Cic. Verr. II 90, senatus et populus Romanus. of the Republic, in Terence, and in Cato. 1 1 (Revue de Philologie, IV i, pp. 25-9), no exception in Sallust, Velleius Paterculus, Q. Curtius, Pliny the younger, Florus, Justin, Cicero (orations), nor in Catullus, Vergil, Ovid, Phaedrus, Persius, Juvenal. In Varro the only exceptions are in quotations. In each of the authors Terence, Horace, and Caesar there is one excep- tion; in Columella, two; Propertius, two; Tibullus, three; Cato, four (beneque four times) ; Nepos, five. The inscriptions, at first sight, seem to violate this rule, but only in the case of the active infinitive of verbs. This seems all the more striking when we consider the aversion elsewhere to attaching it thus to the active infinitive, even in authors who allow it after short e in other classes of words (see Drager, Hist. Synt. II, p. 43). None of the prose writers examined by Harant, above referred to, allow it in this position with the single exception of Cato, and he offers but one instance (de Agri cultura, Keil, p. 98, i, cenareque) ; and the poets overstep this rule only in isolated cases.' The inscriptions, then, presenting as they do the complete reverse of this, are unique in prose writings, as far as research has gone. The only examples they present of que appended to short e (or to a long e that afterwards became short) are as follows : C. 198, 34, (produ)cere proferreque. 199, 29, posidere colereque. 200, 41, gerere habereque. 203, 8, locare mittereque. 203, 9, venire mittereque. 205, XX 17, iudicia dato iudicareque iubeto. 205, XXII 46, proscreibeive veneireque iubeto. We observe that out of the 340 instances of que it is in no case appended to a word ending in e, or in e which afterwards became • e, except when that verb is the active infinitive of a verb. It is interesting to note in this connection that, out of only 213 instances of et (a much smaller number than of que), the word following the conjunction ends in e in 4 cases : C. 198, 26, et unde; C. 198, 27, et unde ; C. 198, 67 ; 577, II 15, lita politaque et cake uda dealbata. In this last case, too, we should expect the last member, like the second, to be added by que. This is, in fact, the only instance I find in the inscriptions of — que et—, while — que-que is of very common occurrence (see §25). Ac occurs only 7 times in the inscriptions, but in two of these is followed by a word ending in • Ter. And. I, 3, 12; Hor. Sat. I, i, 89; Tibullus, I, 3, 34; Plaut, Trin. i, 2, 39; Poen. Prol. 3; Most. 3, 2, 104; Trin. 4, 2, 27. 12 Que, Et, Atque in the Inscriptions e: C. 1007, I, asta ac pellige; C. 1008, 9, in luctu ac solicitudine. Again, wherever two words are connected by que and one of them ends in e (or e, afterward e), the one in e is invariably placed first and que attached to the other word, e. g. C. 198, 11, quaes- tione iudicioque ; C. 198, 19, De nomine deferendo iudicibusque legundeis ; C. 205, XXI 5, sponsione iudicioque ; C. 205, XXI 10, sponsione iudicioque. All this clearly shows a tendency to avoid appending que to e (except in the case of infinitives, where it is the only connective used) by using another conjunction, where possible, or by avoiding the order of words. The. exception which the active infinitive forms to this rule is undoubtedly to be accounted for by supposing that its final e was long in quantity. But it is surprising that no other case occurs of que appended to a word whose final e was originally long but afterward became short (e. g, abl. of 3d declen.). Such an anomaly affords opportunity for conjecture. It may signify that the final e of the active infinitive remained firm in quantity after the e in other classes of words had begun to waver. If this is the true explanation, we should expect to find que appended to the infini- tive most frequently in the oldest writers, and this in fact proves to be the case. Plautus has 4 instances ; Terence a little later has one ; while in the writers of the Augustan age the license had virtually disappeared altogether, the e having at last become fixed as short. The facts above pointed out throw doubt upon a conjectured restoration in C. 198, 78, where the title of a section reads : De provocation — eque danda. Mommsen does not attempt to re- store the original reading here in the Corpus, but Bruns (Pontes luris Romani) has written : De provocation(e immunitat)eque. Wordsworth (Spec. Early Latin) also has this reading, following Bruns. This restoration is very improbable. The — eque be- longs, clearly enough, to an ablative case, but an ablative of the 5th declen. would be more in accordance with the usage else- where. In the Ephemeris Epigraphica, I 3 (p. 7) is preserved an inscrip- tion in which occurs PLEIB PROPAVLQ, where Henzen thinks probaveq. should be read. To this Willmanns remarks : utrum sit probabev(e)qu(e) an probave(t)q(ue) nescimus. The above-men- tioned facts favor the latter. These facts also have an important bearing upon certain emen- dations which have been made in the text of Plautus. In Pseud. of the Republic, in Terence, and in Cato. 13 355 the MSS have : Ego scelestus ni^nc argentum promere possum domo. The editors have deemed it necessary to aher this Une, e. g. Lorenz emends to promere huic ; Ritschl to domo potis sum promere, etc. In True. 2, 4, 74, the MSS have : Non aiides aliquod mihi dare muniisculum. Spengel follows Camerarius and writes dare mihi. The text is also changed by Ritschl, Ribbeck and others, to avoid the ictus on the supposed short e. Similar emendations are made, for the same reason, in Stich. 513 (promit- ter€), Trin. 584 (darg), and elsewhere. It may safely be said that all such changes in the text of Plautus are utterly groundless, in so far as they are based on the supposition that the final e of the infinitive was short. IV. Que vi^iTH Prepositions. The rule that que is usually appended to a noun rather than to a monosyllabic preposition governing it, unless the preposition is repeated (a rule still given in Harpers' Dictionary), has been modi- fied by recent investigation. Klammer remarks' that this rule began to fall into neglect in the time of Livy, and that Seneca never appends que to a monosyllabic preposition even when repeated. Similar remarks regarding Livy as marking the time when que began to be attached to the preposition, even when it had not gone before, are made by Ringe, Zum Sprachgebrauch des Caesar, p. 19 ; Drager, Hist. Synt. II 314, i ; Kiihner, Ausf. Gr. d. lat. Sprache, §113. All these statements are misleading. Ante-classical literature, as Drager remarks, had not been examined in this connection. The inscriptions show that no such rule can hold, for here que is always appended to the preposition, whether monosyllabic or not and whether repeated or not. Ballas shows that this is also the case in Plautus." I find, however, that Terence attaches que to prepositions only three times, and in each case the same preposition has immediately preceded (Haut. 811, cum — cumque ; And. 290, per — perque ; And. 540, per — perque). Cato, on the contrary, attaches que to prepositions twice, but in neither case is the preposition repeated (de Agri cultura, 50, 9, circumque ; 93, 3, proque). In one passage of his orations (Jordan, p. 341, 6) we have : ad Illiberim adque Ruscionem, where Jordan is in doubt whether adque is the conjunction, or preposition with que. Drager (Hist. Synt. p. 35) regards it as the preposition. As there is no ' Animadversiones Annaeanae Grammaticae, p. 40. * De particulis copulativis, p. 19. 14 ■ Que, Et, Atque in the Inscriptions other instance in Cato of que connecting repeated prepositions, while there are 15 instances of et in a similar position and 7 of atque, and since Cato uses atque to connect proper names with about the same frequency as et (see §16), it seems better to take this as the conjunction atque. It would, furthermore, be more in accordance with the general rule of Cato, to make a single prepo- sition govern both substantives; cf. Jordan," p. 15, 1. 8, cum Iphe- genia atque Pylade; 23, i, a recte consulendo atque intellegendo ; 37, 16, inter apparitores atque amicos ; 43, 6, in duritia atque industria ; 45, 10 ; 59, 4 ; 67, 4 ; 67, 5, etc. In the inscriptions we notice several marked peculiarities in the use of et and que with prepositions. The following table shows the prepositions found in connection with conjunctions, the figures representing the number of instances : In que if et'] Ex que I eto Cum I Ad iC?) I De 8 I Ab 2 Extra 6 Inter I Pro 2 Notwithstanding the comparative infrequency of et (occurring only once to que five times), it is still used nearly twice as often as que with the preposition in, even in the laws, where que is so universal.'' On the other hand, de andextra show a strong prefer- ence for que. The other prepositions are not of sufficiently frequent occurrence to warrant any conclusion. In case of a repeated preposition que is used 10 times and et only 6; when not repeated, que 13 and et only 4, showing a stronger preference for que when the preposition is not repeated than when it is. If the old rule were true we should expect the reverse. Atque (ac) is nowhere immediately followed by a preposition. V. Que WITH Relative Pronouns. I notice that in the inscriptions que is commonly appended to the relative pronoun only when a relative has already preceded, e. g. C. 204, I 19, quae — quaeque. Otherwise et is more com- mon, e. g. C. 206, 5, ita utei — et quae; 199, 3, controvosias ' Catonis Fragmenta. '^ All but three of the seven instances of et are from the laws, C. ig8, 58 ; 200, 30; 206, 39 ; 205, XXII 46 ; 551 ; 1028 ; 1418. of the Republic, in Terence, and in Caio. 15 composeiverunt, et qua lege— dixserunt. Compare the following table : 10 > a U3 ,n ^ ^ ►Q ^ ^ -2 C c c c c c c c 2. fi p (D p p c' 3 ^^' D- (tl 3 en . ■ . ,,^. — ro 10 10 10 10 to to to s^ ^ # ^ s Q _to CTi t— t 1— 1 f— t M M t— ( f— ( 00 JO M 1—1 CTi 5. ^ 0\ l-H - 00 s "■■ '* * Cn •^ w 'X) 8^ - 10 Cn CTi ^ as §_ 1 to to cji K) to to 1 X X "^ _00 S. ^ >< ^ 1" i-i w ^ vf 00 1 • XXI 21, 1" 10 M ,_, tti vC ON K} (H M to HH vO vO 4i^ vp ■p 0^ M tt 00 H OJ w w " § C» Cn II ^ 5' ^ !5 &• tr. ^a' •^ >, ^J ■^ M M M to :?*t« 1:1 s Grt ^ 5 *• •».| M M M 10 M !5 >£) OJ «^ vp f- M CTv -J OJ " g U t— 1 u . U\ 1— 1 4^ 1 t— 1 "S. -• 0° 1 10 !S §^ . 1: M ■^ 1 ^ 5". r* ^I M w to t-i 1 6 Que, Et, Atque in the Inscriptions It will be noticed that, when a relative has preceded, que is used 22 times and et only 7. Of these instances of et, 4 occur in the same passage, enumerating distinct things; 2 add relatives in different constructions, in which case there would have been greater danger of ambiguity from the use of que (e. g. 199, 3, qua lege — et qua [adv. =: " where "] ; 206, 35). On the contrary, when no relative has preceded, que is avoided almost entirely, occurring only once, while et is used 7 times. This last is more significant when we consider that all but 2 of the 7 instances are from the laws, where que so greatly predominates. It is probable, as I have already remarked, that this use of et was to avoid the ambiguity arising from such forms as quamque, quaeque, quodque, etc. Of course, where a relative has preceded, there would be but little danger of such ambiguity.' Atque {ac) is not thus used to add a relative clause. No comparison can be made with Cato and Terence in this respect, as they prefer et in every connection, and it accordingly greatly predominates both where the relative clause is, and where it is not, preceded by another relative. Cato : with preceding relative, que 2 times, et 10 " no " o 13 Terence : with preceding relative, 3 10 VI. A Detailed Classification of Que, Et, Atque {Ac'), in THE Inscriptions, in Terence, and in Cato. It is the usual custom in papers of this kind to classify the par- ticles separately, assigning to each one a chapter of its own. Instead of doing this, I have thought best merely to assign one section to each class, bringing together under that section all the instances of all three particles which properly belong there. In doing this I have adopted the following order of treatment : (i) que, et, atque in the inscriptions ; (2) que, et, atque in Terence ; (3) que, et, atque in Cato. By treating of these three styles of writing thus side by side, I have hoped to bring out more clearly the points of cor- respondence and of contrast. In consulting the summaries which are given, one should keep constantly in mind the relative frequency ' I have included in the table only the singidar number of the relative. The above remarks hold true for the plural also, perhaps after the analogy of the singular : quibusque with preceding relative, C. 201, 3 ; 204, II 23 ; 206, 5. of the Republic, in Terence, and in Caio. 17 of each particle in the inscriptions, in Terence, and in Cato. As regards the whole number of conjunctions, Terence and Cato are on an equal footing, each having about the same number of instances of que, el, aique. The inscriptions present only about two-thirds as many instances. These are distributed as follows : Inscriptions : que 340, et 215, atque (ac), 20 Terence: 115 525 276 Cato: 224 529 94 The classification below does not include all the instances occur- ring in polysyndeta. Where there is a series of several terms, each united to the preceding by a conjunction, I have not sepa- rately classified in every case the conjunction of each couplet, as that would have greatly complicated the classification without materially increasing its usefulness. I have, however, cited even these instances where there was anything especially noteworthy or unusual. If, for any purpose, one wishes to collect and similarly classify the conjunctions in polysyndeta, he will find references to all the instances under §25. I have also omitted the instances of connectives which are used only with the last term of a series, these being used merely to close the series without regard to the ideas connected. I. One of the most common uses of que in the inscriptions is to connect two clauses expressing successive acts. The conjunc- tion may here be translated by "and then," "and thereupon." The inscriptions here have que nearly five times as often as both et and afque together. Cato has about the same number of instances of this use as are found in the inscriptions, but he differs from them in a marked degree by using et nearly twice as often as both que and atque together. Terence has only one-fifth as many instances of this use as the inscriptions, or Cato, and at the same time presents a further contrast by using most frequently atque, the particle which, in the inscriptions and in Cato, appears least frequently. Compare the following : Inscriptions. Que, Future Imperative: C. 198, 7 [praetor recuperatores n. n. dato] ' deque eo homine — [ioudicare iubeto] ; C. 198, 18 ; 198, 19, in ious educito nomenque eius deferto ; C. 198, 20, 59, 64 ; 200, 39, 78, 84 ; 202, 1 I ; 202, II 20 (twice) ; 209, 6; 577,11 16; 577,111 i; Perf. Indicative: C. 195, 8; 551, 11; ' Conjectured restorations are enclosed in brackets. The numbers refer to the first volume of the Corpus Inscriptionum Latinavum. 1 8 Que, Et, Atque in the Inscriptions 1252, 5; the formula, faciund. coiravit eidemque probavit (with variations of number, etc.), occurs 31 times : C. 566, 567, 576, 591, 594, 600, 605, 802, 1140, 1149, 1161, 1163, 1178, 1187, 1189, 1192, 1216, 1223, 1227, 1245, 1247, 1250, 1251, 1259, 1279, 1341, 1407, 1421, and in 1150, prob. dedicarq. ; Subj. : C. 196, 48; 196, 17, 27 ; 205, XX 21, 52 ; 205, XXI 24 ; 208, 3 ; Inf. : C. 199, 4 ; 200, 97 ; 205, XXII 45 ; Fut. Perf. Ind. : 198, 24 ; que sometimes con- nects participles that may, not improperly, be classified under this section: C. 198, 18, descriptos habeto eosque propositos, etc.; 198, 38; 577, II 18, lita politaque — facito (shall have them smeared over and then polished) ; 577, III 15. Et, Fut. Imp. : 197, 17, 21; 198, 53,58; 577, II 10; Perf. Ind.: 199, i; 551, 1028, 1135, 1258 ; Subj. : 203, 21 (the only instance), [in integram restitu]antur et de integro iudicium — fiat ; Fut. Perf. Ind. : 198, 32, 76; Gerundives : 1196. Atque, thus used but once : 196, 28. Terence, ^m^ occurs but once in this use : Hec. 508, deliberet renuntietque. Et, Perf Ind. : Phorm. 1006, uxorem duxit et inde filiam suscepit ; Hec. 376; Subj.: Adelph. 316, 446; Inf.: Phorm. 150; Pres. Ind.: Haut. 651. Atque, Perf. Ind.: Haut. Ill, Hec. 169, Haut. 144 {ac); Subj.: Phorm. 586 ; Inf : Phorm. 414; Fut. Ind.: Adelph. 181; Gerundives: Haut. 509; Pres. Imperat. : And. 725. Cato. Cato's de Agri cultura is, as far as the form of compo- sition is concerned, very similar to legal documents — that is, it consists largely of directions as to what shall and what shall not be done. There are, accordingly, a large number of future impera- tives, as in the inscriptions. Cato, however, more commonly connects hy et. Que occurs as follows: de Agr. cult. p. 31, 20;" 36, 11; 38, 8 ; 3S, 18 ; 39, I ; 41, 2, 4 (twice) ; 44, 12 ; 48, 6 ; 50, 10 ; 51, 13 ; 52, 12, 20 ; 60, 5, 18 (twice) ; 61, i, 13, 15 ; 66, 13, 16 ; 67, i ; 81, 12; 82, 5, 8; 95, 17; Perf. Ind.: Fragm. p. 56, 3;' Subj.: de Agr. cult. p. 65, 2; 98, 15; Inf : de Agr. cult. p. 53. 9; Parti- ciples : Fragm. p. 21, 7, proelium factum depugnatumque. Et, Fut. Imp.: de Agr. cult. p. 18,9, 10; 36,7; 41,6; 43, 11; 48,19; 53> 6; 53, 18; 59, 5; 61, I ; 62, 6; 65, 9; 66, 10; 68, 5; 71, 17; 73, I ; 74, 2, 4 ; 75, I, ID ; 76, 3 ; 77, 9, 13 ; 78, 7 ; 79, 4 ; 80, 7, 10; 98, 2; 104, 20; 105, 7 ; 107, II ; Perf Ind. not found; Subj.: p. 34, 12; 60, 5; 76,11; Inf: p. 15, 12; 53, 10; 84, 6; 96, 2; 104, 3 ; Pres. Ind. : p. 19, 13 ; 52, 17 ; Fut. Ind. : p. 76, 5 ; 108, ' The pages refer to Keil's edition of de Agri cultura (Leipzig, 18-82), and to Jordan's edition of the fragments (Leipzig, 1862). of the Republic, in Terence, and in Cato. 19 3 ; Fut. Pf. : p. 37, I ; 74, 2 ; Gerundives : p. 91, 5, 7 ; Fragm. p. 43, 8; Pres. Imperat. : de Agr. cult. p. 44, 7 ; 107, i. Atgue, only once : Fragm. p. 25, 10, exercitum — eduxit foras atque instruxit. 2. Rarely the act expressed by the second clause refers to time prior to that of the preceding. Inscriptions. Que: C. 201, 5, ea senatus animum advortit ita utei aequom fuit. nosque ea ita audiveramus, ut vos deixsistis ; 206, 35, earn viam — tuendam locato. isque aedilis — propositum habeto quam viam tuendam et quo die locaturus sit ; 206, 19 ; 206, 97, 107, 125, 140; 207; 1409, 5. Et and atque not found. Terence. Only once : And. 836, facta atque incepta omnia. Cato. Que : de Agr. cult. p. 37, 2, facito uti servetur — facitoque studeas bene percoctum siccumque legere. Et : de Agr. cult. p. 29, 4; Fragm. p. 72, i. 3. Que and et, and, in a few passages in Terence, atque, are used to subjoin a clause more accurately defining or explaining what has gone before. Sometimes the added clause merely specializes or adds a less comprehensive notion. Terence presents the largest number of instances of this use, having considerably more than both the inscriptions and Cato combined. Inscriptions. Que : C. 204, 34, ea omnia — utei sunt fuerunt, ita sunto itemque ieis ea omnia habere possidere uutei frueique liceto (everything shall be as it was, namely, they shall be allowed, etc.) ; 205, XXI 15 ; 205, XXII 45, 47 ; 206, 146, colonorum — censum agito eorumque nomina praenomina patres — accipito; 577,1 7; 1422; 195, 11; 201, 3. Et: 199, 3, contro- vosias composeiverunt et qua lege agrum possiderent — dixserunt; 1009, 13. Terence. Que: And. 556, 649; Phorm. 549; Hec. 207, 478, 579i 755 ; And. 935. Et, used much more frequently : And. 22 ; Eun. 1087 ; Haut. 159, 424, 735,962, 1034; Phorm. 54, 381 ; Hec. 25, 268, 315; Adelph. 272, 694 ; And. 97; Phorm. 290 ; Hec. 117, 376; Adelph. 2. Atque : And. 15, 627; Adelph. 217, 846, 980; Eun. 385. Ac : Haut. 728; Eun. 92. ■ Cato. Que : de Agr. cult. p. 52, 2, depurgato herbasque malas omnis radicitus effodito; 22, 3; 50, 16; 62, 11; 93, 3; Fragm. p. 19, 13. Et : de Agr. cult. p. 40, 8; loi, 19; 28, 8; 29, I ; 41, 9 ; 57, 8 ; 102, i ; Fragm. p. 77, 8. Atque, only twice, i)oth times in the Fragments, p. 20, .2 ; 88, 2. 20 Que, Et, At que in the Inscriptions 4. Opposed to the examples in the preceding section is a class in which the added clause is of a generalizing character, or expresses a notion which includes that of the preceding clause. The con- junction may often be translated by "and in general." There is but one example of this use in the inscriptions. It is not uncom- mon in Terence and in Cato, the former using et or atque {que only once) ; the latter, only que and et. Inscriptions : C. 206, 78, scaenam pulpitum — in loco publico ponere statuere eisque diebus — loco publico utei liceat. Here the Using of the place includes the erection of the scaena, etc. Terence. Que : Hec. 604. Et : Haut. 7, novam esse ostendi et quae esset; Phorm. 512, 759; Adelph. 429. Atque: Adelph. 33. 794. 880. Ac : Hec. 654. Cato. ' Que : de Agr. cult. p. 13, 10, quo modo fundus cultus siet operaque quae facta infectaque sunt ; p. 47, 11 ; 83, 5 ; 89, 2. Et : de Agr. cult. p. 28, 3, materiem et quae opus sunt ; 40, 7 ; 82, 17. 5. A copulative particle is sometimes used to connect clauses of which the sfecond bears an adversative relation to the preceding. In such cases we should be more likely in English to use an adver- sative conjunction and, in translating, we may accordingly render by "but," " and still." Et is by far the most common particle in this connection. In the inscriptions et is thus used three times, que once, atque not at all. In Terence, on the other hand, atque is thus used with equal freedom (et 29, atque 23, que 2). Cato, again, uses et almost exclusively (et 15, atque 3, que i). I may note here, in passing, that Schmalz is wrong when he says (Latein- ische Grammatik, §166) that Cicero was the first to use neque in this adversative sense. It has that force distinctly in C. 205, XXI 15, quel ita quid confessus erit neque id solvet (who shall have confessed his debt, but shall not pay it); also Terence, Hec. 641, Haut. 982, and elsewhere. (a) First to be noticed under this section is a class in which the added clause is negatived by non, nunquam, or the like. Inscriptions. There is, strictly speaking, no example of this use in the inscriptions, but the following may be cited as having virtually, in the second clause, a negative force : C. 692, esureis et me celas (you are hungry, but will not tell me so) ; 199, 45 (where abstineant really expresses a negative idea). Terence. Que not used. Et : Eun. 76, Haut. 387, Phorm. of the Republic, in Terence, and in Cato. 21 521, poUicitantem et nihil ferentem. Atque not thus used, but in the following passage the added clause expresses a negative notion : And. 299. Ac: Eun. 451, And. 873. Cato. Et is here the only connective employed : de Agr. cult. P- 12, 2 ; 75, 4 ; 105, 5. {U) Sometimes Xk\.z first clause is negatived and the second adds an affirmative notion adversative to it. The most common case of this kind is where both clauses are imperatives. Inscriptions. Only once: C. 205, XIX 2 ique). Terence. Que not used. Et: Eun. 78, And. 59, Eun. 16, ne erret moveo et desinat lacessere; Eun. 965, 1075 ; Haut. 49, 176 ; Hec. 50, 220, 487, 561 ; negative in force: Phorm. 857. Atque: Haut. 84, And. 225, 614 ; Eun. 51 ; first clause with negative force : Adelph. 185, 335, 755. Cato. Only et used : de Agr. cult. p. 67, 16, scabrae non fient et multo feraciores erunt; p. 15, 1.4 ; 69, 5 ; 80, 12; loi, 20. (<:) Again, the particle may join two affirmative clauses which seem inconsistent with each other, or even opposed to each other. In the inscriptions and in Cato this use is very rare and the con- nective is invariably et ; in Terence there are numerous examples, the connective here also being generally et, sometimes que or atque. Inscriptions. Only once : C. 206, 157, qui pluribus in muni- cipieis — domicilium habebit et is Romae census erit. Terence. Que: Hec. 56, 199. Et : Eun. 2, Haut. 663, 696, abis et Bacchidem hie relinquis ; Haut. 867 ; Phorm. 1053, Hec. 344, 347, 401,610, 792; Eun. 481. Atque : And. 339, Hec. 844. Cato. Only et : de Agr. cult. p. 44, 7 ; 103, 9. {d') Often in Terence atque is used to introduce a clause which is adversative to a remark just made by another character, or to some reflection of the speaker himself. Sometimes it merely intro- duces something wholly unexpected and is followed by eccum. There is no example of this use in the inscriptions or in Cato. Ballas finds that Plautus uses both atque. and et with eccum. Terence has only atque, Terence. Atque : And. 350, 607, ubi illic est ? scelus qui me hodie — atque hoc confiteor iure mihi qptigisse; And. 640, 977; Eun. 480, Haut. 187, 686; Adelph. 362. Ac: Phorm. 232, Adelph. 626, And. 370. Alque eccum (generally said at the unex- pected appearance of some person) : And. 532, 579, evocate hue Davoni. Atque eccum video ipsum foras exire ; Arid. 579, 957 ; Eun. 455, 1005; Hec. 246, 352, 523. 22 Que, Et, Aique in the Inscriptions {e) Et and aique (never que) are used in Terence and Cato to add a notion which is true, in spite of a preceding statement with which it seems inconsistent. It may be rendered, in translating, by "and still," " but still." Often this relation is expressed by the addition of tamen. Inscriptions. No example. Terence. Et : Eun. 72, Adelph. 596, 726, scis et patere ? Et tamen : And. 633, timent et tamen res premit ; Eun. 24, Haut. 567, 933 ; Hec. 465. Atque : Haut. 195, Phorm. 389, Adelph. 40. Aique tamen : Haut. 205. Cato, only in the Fragments. Et : p. 24, 13. Et tamen: p. lo.^, I ; 107, 2. Atque : Fragm. p. 24, 12. Aique tamen: p. 23, 12. (/) In two passages in Cato the added clause constitutes an exception to the preceding. Que : de Agr. cult. p. 52, 11. Et : p. 102, 13. 6. Two clauses of similar meaning are often joined for judicial fullness, or to fill out and round off a single complete notion A marked difference is to be noted here in the use of the conjunctions. In the inscriptions such clauses are almost invariably joined by que {que 19, et 2, atque 6). In Terence, on the other hand, et is by far the most frequent {et 12, que 3, atque 3, ac 2) ; while in Cato the three conjunctions are used indiscriminately in this connec- tion {que 8, et 5, atque 5). Inscriptions. Que : C. 203, 8 ; 206, 26, 30, 52 ; 1008 ; 198, 26, 29, 69; 199, 24, possidere fruique ; 199, 30, 44, mittei lerbera- reique; 200, 16, 24, 28, 87, 89; 202, II 25; 603, 7; 1027, 1408. Et: C. 1097, 17; 1408. Terence. Que: Eun. 84, Hec. 388, tecta tacitaque aput omnes sient ; Adelph. 392. Et : And. 296, 648 ; Eun. 103, 886 ; Haut. 260, 418, 425, 926,966; Adelph. 68, 994. Atque: Hec. 297, 686; And. 809. Ac : And. 62. Cato. Que : de Agr. cult. p. 49, 10 ; 85, 9 ; 87, 10, 18 ; 88, 12 ; 89, 4; 100, 15 ; loi, I. Et: de Agr. cult. p. 28, 3 ; 35, 9 ; 100, 5; ic6, 10; Fragm. p. 9, 10. Atque, only in the Fragments: p. 22, I, 2 ; 36, I ; 55, 8 ; 73, 2, uti atque frui. 7. Que and atque are sometimes, and et very often, used to add a clause, the relation of which to the preceding may be expressed in translation by " and so," " and for that reason." This of the Republic, in Terence, and in Calo. 23 use is especially frequent in Terence ; in Cato it is comparatively rare, and in the inscriptions still less frequent. Inscriptions. Que: C. 198, 9 ; 205, XIX 2 ; 206,127. ^^ ■' C. 1006, bene rem geras et valeas dormias sine qura ; 1202 ; 1253 ; 201, 9 ; 1 192. Terence. Que: And. 488, quandoquidem ipsest ingenio bono cumque huic est veritus optumae adulescenti facere iniuriam (of a good disposition and so shrank from injuring, etc.) ; And. 585, Haut. 1059, Phorm. 843, Hec. 203, 386, 396. Often an additional word or phrase is used to bring out the relation between the clauses : itaque : And. 550, Eun. 317, 945; Phorm. 870, Hec. 201,802; Adelph. 258, 7 ID ; proptereaque : And. 693 ; que with ob earn rem : Hec. 749. Et: And. 51, 577 ; Eun. 7, 54, 72, 260, 384, 464 ; Haut. 197, 244, 434, 504, 703, 1031 ; Phorm. 2, 69, 127, 316, 405, 452, 456, 886; Hec. 55, 264, 748; Adelph. 107,138,380,680,729,886; et proplerea : And. 653; Eun. 879; Hec. 871 ; et ob earn rem: Adelph. 895; etea gratia: And. 433. Atque : And. 525 ; Haut. 535, 860, 1026; Phorm. 746, 894; Adelph. 283. Ac: Adelph. 624; atque ita : Phorm. 716. Cato. Que : de Agr. cult. p. 61, 11 ; 87, 15 ; 98, 3 ; Fragm. p. 19, 10. Et: de Agr. cult. p. 102, 2; 103, 18; Fragm. p. 77, 7. Atque : de Agr. cult. p. 35, i ; 23, i ; 51, i. 8. Sometimes the added clause expresses the cause of, or the reason for, the statement in the preceding clause. The relation between such clauses may, accordingly, often be brought out by translating " for the reason that," " inasmuch as." This use occurs only once in the inscriptions, and is rare in Cato ; but instances of it in Terence are not uncommon. The connective is generally et. Inscriptions. Et : C. 1009, 19, rehqui fletum nata genitori meo et antecessi leti diem. Terence. Que : Eun. 333 ; Haut. 168, 445 ; Phorm. 480. Et: Haut. 105, 116, 259, 381 ; Phorm. 721 ; Adelph. 64, 121, 580. Atque: Eun. 198; Haut. 734 ; Phorm. 323. Ac: Phorm. 648. Cato. Que : de Agr, cult. p. 99, 14. Et: de Agr. cult. p. 69, 7 ; 79> I- 9. Et is often used in Terence in the sense of " and too," " and likewise." It is most frequently found in connection with the pro- nouns tu and ego, and generally introduces a reply to the remark 24 Que, Et, Aique in the Inscriptions of some other person. Somedmes it is combined with a word like quogue, etiam, item, etc. Que and atque both occur in this use, but are very rare. Only two instances have been noted in the inscriptions, and only five or six in Cato (counting all three particles). Inscriptions. Only que, and only in connection with item : C. 198, 52 ; 204, I 34. Terence. Only et thus used : Eun. 191, in hoc biduom Thais vale— mi Phaedria, et tu ; And. 347 ; Haut. 167, 739 ; Phorm. 209; Hec. 83, 194, 197, 606; Adelph. 129, 751, 972, gaudeo — et ego; with item: Adelph. 230; una: Haut. 191; Adelph. 753; simul: Haut. 803, 943 ; Hec. 792. Atque occurs nowhere alone, but with qiioque in Phorm. 877 ; and with etiam in Adelph. 209. Cato. The only examples noticed are in connection with the particles item and simul; itemque : de Agr. cult. p. 63, 11; et item: de Agr. cult. 11, 5; 64, i; 98, 2; 100, 11; et simul: p. 68, 5. It may not be out of place here to call attention to an error of Schmalz. In his Lateinische Syntax, §169, he cites Adelph. 129 (curae est mihi — et mihi curae est) as an instance where Terence uses et in the sense of " also." This use of et is, however, quite distinct from that with which Schmalz wishes to identify it, as, for instance, in the example cited by him : addito et oleum. It is exactly parallel to our "and too" (e. g. "and me too"), f^ here acting the part of an introductory and transitional particle, in addi- tion to its function of adding.' There is no instance in Terence where et is used in the sense of " also," pure and simple, and the paragraph in Schmalz should be corrected accordingly. 10. Que and et in Terence and et in Cato are often used in making a transition to a new subject or thought. Sometimes a transitional word or clause is added to foreshadow the nature of what follows. Que alone is thus used in the inscriptions, and but twice. Que and atque are also found in a few passages in Cato. Inscriptions. Que: C. 196, 23; 201, 12, quomque de eieis rebus senatuei purgati estis, credimus, etc. Terence. Que: Haut. 383, 525, 585 ; Phorm. 763, 898 ; Hec. 208, 490, 858; with ^-clause: Hec. 471; with quod: Hec. 581. ' This remark is not inconsistent with the views expressed under the second division of this paper, as I have there restricted my remarks to particular uses. of the Republic, in Terence, and in Cato. 25 Et: Haut. 239, 248, 556, 705, 775, 786, 854; Phorm. 471 ; Hec. 523; with quod: Haut. 204; Eun. 64; with quia: Eun. 586; And. 122; with quantum: Haut. 984; Hec. 460; et nunc : And. 157. Atque: Adelph. 850. Ac: Haut. 948. Cato. Que: Fragm. p. 42, 9; 55, 8. Et : de Agr. cult. p. 22, 7 ; 70, 3; 100, II, 13; with quoin: p. 11, 8; with uti : p. 50, 7 ; with si : p. 67, 11, et idem hoc si facies ad arbores ferace.s, eae quoque meliores fient ; p. 69, 4; 70, 3; 99, 10; loi, 7, 9, 10, 12, 15, 16; 102, 3; 103, 10, 17; 104, 13, 15, 19; 105, 16. Atque: Fragm. p. 23, 4, 7 ; 62, 3 ; 63, 6 ; 77, 4. II. Et and, less frequently, que and atque serve to unite two clauses of which the second corrects the first, either by the use of a more suitable expression, or by adding a more emphatic or sur- prisingly comprehensive notion. In such cases the relation between the clauses may be brought out by translating the connective by "or rather," "or perhaps I should say," "and what is more." This use does not ocCur with any freedom except in Terence. The connective is usually et. Inscriptions. Que: C. iy8, 18 [permittito potestatem]que scribundi, quei volet, facito (shall permit him to write and, what is more, shall provide the means). Et : C. 205, XX 11. Atque : C. 1008, 12, oro atque obsecro. Terence. Que: Phorm. 866 ; And. 592. Et: Eun. 72, 744; Haut. Prol. 19; Phorm. 1006; Hec. 166, 241, 615; Adelph. 207, 389, 521, 648, ut opinor et certo scio ; Adelph. 964. Atque: And. 823.' Cato. Que : Fragm. p. 23, 2. Et:de Agr. cult. p. 11, 2; 75. 5 ; 104, 7 ; 67, 14 ; 103, 11. Atque: Fragm. p. 36, 15. 12. Not infrequently in Latin a clause is added to express the purpose of an act just specified. This is especially frequent with clauses of command, where the verbs are in the imperative mood, e. g. abi atque eum require. Ordinarily such imperatives stand without any connective (Drager, Hist. Synt. II, p. 27), and Schmalz (Lat. Synt. §163 Anm.) claims that this was, without doubt, the original construction. If this be true, we would naturally expect the fewest instances of a connective in the oldest writings, and in writings which represent the language of the lower classes ; for it is among ' For a similar use of atque with adjectives and adverbs, see under §§22, 23. 26 Que, Et, Atque in the Inscriptions the lower classes that primitive expressions and constructions are longest preserved. We are, however, confronted by the fact that the comic poets, who introduce so much from the sermo vulgaris, are just the ones to employ the connective freely, while most of the later writers, with more polished diction, regularly have asyn- deton, e. g. Livy and Vergil, while Horace, Ovid, Martial, and Seneca use the connective only when it is accompanied by nunc (Schraalz, Lat. Synt. p. 301). Ballas cites a large number of instances of a connective from Plautus (Gram. Plaut. I, p. 15 ff.). I have collected below numerous instances from Terence. In the inscriptions, as far as I have noticed, such imperatives always have a connective (et or ac). Perhaps these observations are not suffi- ciently extended to warrant my contradicting the statement of Schmalz that, historically, abi require preceded abi et require. But there is certainly nothing in Plautus, Terence, or the inscrip- tions to justify such a statement ; on the contrary, there is much to favor the view that the use of a connective with such imperatives lays claim to equally remote antiquity as asyndeton. With respect to the conjunctions employed to add a purpose clause, the inscriptions, Terence, and Cato differ in a marked degree. The inscriptions almost invariably use et. Terence, with but very few exceptions, uses atque {ac) ; and Cato, with still fewer exceptions, uses que. Inscriptions. Que, Imperatives : no example ; Subjunct. : C. 199,34, materiam sumant utanturque ; 206, 153; Infin. : 198, 34. Et, Imperat. : 1009, 4, morare gressum et titulum nostrum perlege; 1027; 1306; Subjunct.: 819; Perf. Ind. : 1009,8. Ac, Imperat. : 1007, i. Terence. Que, Imperatives : no example ; Subjunct. and Inf : no example ; Pres. Indie. : Hec. 145 ; Gerund : Hec. 92, abeundi vosque videndi. Et, Imperat. after abi : Phorm. 564 ; Adelph. 917 ; after i: Hec. 611 ; Adelph. 854 ; Subjunct. : Eun. 808, 1026 ; And. 639 ; Inf. : Eun. 467 ; Phorm. 252, adire et blandi adloqui ; Pres. Ind.: Haut. 1048; Hec. 430; Fut. Ind.: Adelph. 591 (?). Atque, Imperat. after abi: Eun. 763; Haut. 619; Phorm. 309; Adelph. 351. Ac: And. 255; Hec. 314 ; Adelph. 168,699; other verbs : And. 727 ; Phorm. 921 ; Hec. 359. Ac : Haut. 831 ; Subjunct. : And. 542 ; Phorm. 881 ; Hec. 754 ; Adelph. 300, 599, 453.786; Infin.: And. 14; Adelph. 416; Pres. Ind. : Phorm. 845. Ac : Adelph. 916; Fut. Ind. : And. 599 ; Eun. 206, 216, 557, 922 ; Phorm. 312; Adelph. 513, 590; Hec. 515. Ac: Adelph. 510. of the Republic, in Terence, and in Caio. 27 In Hec. 285 the infinitive after atque expresses rather the result : redire atque resciscere (return, only to find). Cato. Que, Fut. Imperat. : only que used and that, too, when et greatly predominates in Cato : de Agr. cult. p. 36, 15 ; 51,5; 61, 17, 18; 83, 5, 6; Subjunct. : p. 89, i ; 98, 16. Et, Subjunct. : de Agr. cult. p. 80, 6 ; Inf. : 97, 8. Atque, Inf : Fragm. p. 54, 8. 13. There is sometimes a union of two clauses between which no special relation suggests itself. They are simply distinct, co-ordinate statements. Et is naturally most frequent here except in the inscriptions, where que greatly predominates. Inscriptions. Que: C. 200, 8,46, mancepspraevidespraediaque soluti sunto ; eaque nomina — in tableis [publiceis scripta habeto ; 200, 66 ; 202, In; 202, II 10 ; 203, 8 ; 204, 1 8 (twice) ; 204, 1 19 ; 206, 35, 44 ; 571 ; 1008, 6 ; 195, 5 ; 551, 12 ; 1474. Et .- C. 551, 9; 565; 1012; 1019,7; 1246,5. Atque: C. 196, 26. Sometimes the clauses are not merely distinct, but contrasted as well. Que : C. 197, 21 ; 198, 21 ; 202, I 6 ; 205, XXII 51 ; 206, 21 ; 1166, 15. Et: C. 1 25 1. Terence. Que: Hec. 146. ^i".- And. 42, 66, 175, 498, 515, 684 ; Eun. 69, 133, 160, 203, 215, 419, 492, 513, 572, 795, 1094; Haut. 304, 306, 1040, 1067 ; Phorm. 189, 1000, 1055 ; Hec. 88, 225,599; Adelph. 43, 258, 879. Atque: And. 268 ; Eun. 10,461, 588; Haut. 900; Adelph. 154. Ac: Haut. 633. Contrasted. Que : And. 777. Et : Haut. 152 ; Phorm. 126 ; Adelph. 352, 571. Ac : Haut. 927. Cato. Que: de Agr. cult. p. 12, 9; 40, 6 ; 49, 10; 58, 12- 63, 3; 87, 14; 98, I. Et: p. 15, 13; 17, 21 ; 19, 3, 13; 40, 12; 41, 6; 45, 6; 52, 4; 70, 2; 75, 9; 99, 7 ; loi, 5 ; Fragm. p. 22, 5. Atque: Fragm. p. 35, 4, 10; 58, i. Contrasted. Que : de Agr. cult. p. 100,4. ^i- P- 17. 18; 39, 10; 52, 5; 66,4; 83, 7; 96, 14. Atque : Fragm. p. 42, 2. 14. Que in the inscriptions, et in Terence, and both et and que in Cato, are often used to add a notion that is merely acces- sory to the preceding or modificatory of it. The connective may be translated by "and that too." Inscriptions. Que : C. 197, 12, sei quis magistratus multam inrogare volet, liceto eique omnium rerum siremps lexs esto (it shall be allowed him, and that too under the same conditions of law, etc.) ; 202, 1 39 ; 205, XXII 47 ; sometimes a pronoun is used 28 Que, Et, Atque in the Inscriptions referring to the preceding clause : C. 205, XX 9, ab eo quei ibei iure deicundo [praerit] postulaverit idque non kalumniae kaussa, etc. ; 202 I 4, 32 ; 204, I 23 ; 206, 37 ; 1230. Et : 204, II 30. Terence. Que: Haut. 368. Et : And. 511; Haut. 114, 302, 726; Phorm. 716, 739; Hec. 239, 402, 692; Adelph. 30, 811, 842, 888, 897 ; with pronoun expressed : Haut. 934 ; Hec. iii. In the following, translate by ''and not, either" the negative of "and that too": Haut. 248, vesperascit, et non noverunt viam (it is getting late and they do not know the way either) ; Phorm. 104 ; Adelph. 43, 122. Atque: Eun. 956. Ac : Phorm. 275; pronoun expressed: Haut. 461: Phorm. 197 ; And. 692 ; " and not either " {ac): Haut. 999. 15. Sometimes the parts connected have to each other the relation of alternatives, which we should be more likely to connect with " or," ■" or, as the case may be." Que is, with the exception of one passage, the only connective thus used in the inscriptions ; et is the regular connective in Cato ; I have noted but one example of this use in Terence. Inscriptions. Que: C. 200, 84; 203, 27, legatos venire mittereque liceret (to come themselves as ambassadors, or to send others in that capacity — whichever they might choose) ; 202, H 35 ; 199, 5, 36 ; Substantives : C. 202, 1 3, ei scribae scribeisque heredive eius solvito ; 206, 11, 68, 72; Pronouns: 2o5, 38. Et : 198- 34- Terence. Que: And. 214. Et : Eun. 258. Cato. Que, only twice and with adjectives : de Agr. cult. p. 77, 4, 5. Et: de Agr. cult. p. 40, 10 ; 69, 15 ; 79, 9 ; 102, 3 ; Sub* stantives : p. 15, 15 (cf p. 57, 10) ; 69, 14 ; loi, 15 ; Prep, phrases: 57, 4; "or, in other words": p. 49, i. Atque: de Agr. cult. p. 57. 18 (cf- P- 57. 14)- 16. The foregoing sections have occupied themselves chiefly with clauses. We come now to the use of the copulative con- junctions with substantives. And under this head let us consider them, first, as connecting two species of a single genus, where a third term is at once suggested which comprehends them both. For instance, when we read ursi et leones, we have suggested to us at once a third term, animal. (a) In general. Inscriptions. Que : C. 198, 73, praetori quaestorique ; 199, 34; 200, 95; 202, 30, 35; 577, II 16; 635; of the Republic, in Terence, and in Caio, 29 I158; I2yi. Et: 577, II 10; 804; looS, 12; 1089; 1413; 206, 10,50. Alque : 196,19; 542. Ac; 1012,7. Terence. Que only once (cf. inscriptions, where it is the most common connective in such cases) : And. 161. Ei : Eun. 26, iii, 112, 518, 840; Adelph. 747, 905. Alque: Haut. 223; Hec. 815; Eun. 746. Cato. Qtie: de Agr. cult. p. 13, 17 ; 18, 8; 37, 2; 40, 5 ; 43, 8 ; 52, 3. 13 ; 88, 6; 89, 7 ; Fragm. p. 30, i. Et : de Agr. cult. p. 19, 7, 8, 10 ; 27, 10 ; 42, 6 ; 44, 6, 7 ; 45, 8 ; 47, 14 ; 48, 9, 16 ; 49> 2 ; 54, 8 ; 55, 3 ; 58, 18 ; 61, 5 ; 66, 1,9; 71,8; 72. 4 ; 76, 13 ; 79. 5 ; 80, 10 ; 81, 10, 15 ; 85, i, 3, 5 (twice), 8 ; 88, 3 ; 90, 3 ; 93, i ; 94, 12 ; 105, 12 ; 109, 8 ; Fragm. p. 64, 2. Alque : de Agr. cult. P- 39. 15; Fragm. p. 37, 16; 45, 10; 56, i ; 81, 10. Ac: de Agr. cult. p. 47, 13. (Ji) According to Drager (Hist. Synt. II 44), the masculine form of the plural of a substantive is connected with the feminine form of the same substantive generally by que. Ballas (Gram. Plaut. I, p. 25) cites 24 instances of this use from Plautus, of which all but one have the connective que. The inscriptions, on the contrary, have only et, never que. Inscriptions. Et : C. 1059, liberteis et libertabus ; 1063; 1253. Terence. Que : Phorm. 976 ; Hec. 102 ; Eun. 302 (in each case di deaeque). Et. not used. Alque not used, though senex atque anus occurs in Hec. 621. Cato has no examples of this. {c) Proper names. Inscriptions. Que : C. 32, Corsica Aleriaque; 197, 17, 25; 199, 33; 547. Et : 199, i, 14, 23; 204, 15 (twice); loio; 1199 (twice); 1313 (twice); 547; 548; 549; 567; 569; 589; 619; 776b; 1024; 1432; 1034; 1036; 1055; 1183 ; 1217 ; 1241 ; 1479; Personal Pronouns and proper name : 1276; 1433. Terence. Only et used : Haut. 498 ; Eun. 25, 732 ; second name preceded by tu: Hec. 664, Laches et tu Pamphile; Eun. 1086 ; only one of the substantives a proper name : And. 552, 924 ; Phorm. 218, 1036 ; Hec. 449. Cato. Que : de Agr. cult. p. 87, 9. Et :■ p. 55, 2 ; 83, 9, 11 ; Fragm. p. 11, 17. Atque: Fragm. p. 15, 8; 15, 7 J 34. 6. 17. In the following the words connected do not suggest a common class, but, rather, different classes. The most common connective is, in each case, et. ri 30 Que, Et, Atque in the Inscriptions Inscriptions, jg'wf .■ C. 196, 22 ; 198, 19, de nomine deferendo indicibusque legundeis ; 198, 78(?); 551; 1071; 1161. Et : 205, XX 42, ea nomina et municipium ; 541 ; 577, I 15 ; 592; 1199; 1418; 564; 570; 1140; 1181 ; 1307; 1421. Terence. Que: Phorm. 890; Hec. 404; Eun. 236. Et : Haut. 479, 486; Phorm. 1012; Hec. 75; Adelph. 57, 230, 495, 585, 847; And. 34, 2S8, 369; Eun. 265, 375, 778, 941. Atque: Haut. 452, 778 ; Phorm. 1024 ; And. 72, 286, 855, 880. Cato. Que : de Agr. cult. p. 12, 14; 38, 12; 60, 16; 87, 16 92, 14; Fragm. p. 25, 6 ; 28, 13 ; 35, 8. Et : de Agr. cult. p. 12 15; 24, 12; 27, 13; 30, 7; 31, 8, 13, 15, 17; 34,8, II, 17; 35, 5 37, 2; 47, I ; 57, 14, 20; 58, I ; 61, 12; 64, 12; 68, 3, 7, 11 ; 79 7 ; 83, 8 ; 85, 2, 3, 7 ; 89, 16 ; 92, 7 ; 94, i, 17 ; 98, 12 ; 99, 7, 8 102, 12; 104, 10; 108, 14; Fragm. p. 39, 11. Atque: Fragm. P- 59. 4 ; 67, 4, 5. 18. Often the conjunction adds something that belongs to the preceding or accompanies it. The second member, accord- ingly, generally has the genitive of a pronoun depending upon it and referring to the first member. Inscriptions. Que : C. 198, 60, eiei iudicei consilioque eius; 204,6; 1195; 1418; 203,9; 1059; 1065. Et: 206, 43, the only instance. It is, however, the only particle used to add the pos- sessive pronoun to the personal : C. 1023, sibi et sueis ; 1042 ; 1070; 1 1 80; 1208; 1244 (twice); 1271; 1460; 1429; Substantive expressed: 1056, sibei et suis liberteis; 1207; 1223; 1242. A strange combination oiet — que is found in C. 1041, sibi et sueis^a^; also 1229. Atque not found. » Terence, jg^we not used. Et: Haut. 961 ; Phorm. 217; And. 538. Atque : Haut. 455. Cato. jg'wif .• de Agr. cult. p. 82, 16, 18. ^/.-p. 89, 15. Atque: Fragm. p. 35, 10. 19. Substantives of cognate meaning are sometimes joined to fill out and complete a single comprehensive notion. The par- ticles seem to be used here indiscriminately. Inscriptions. Qtie .• C. 198, 11, quaestione iudicioque publico ; 202, I 35; II I ; 547; 548; 549. Et: 577, III 5; 566; 567; 568; 1220. Ac: 1008,9. Terence. £'«^.- Hec. 48,fautrix adiutrixque; And. 114; Eun. 300, 815. El: Haut. 710, 945 ; Phorm. 441, 473 ; Hec. 2, 43, 797 ; of the Republic, in Terence, and in. Caio. 31 Adelph. 391 ; Eun. 932, 1048. Atque: ITaut. 417,490; Hec. 165, 123, 860; Adelph. 297, 829, 869; And. 880, 200, 831 ; Eun. 234. Ac : Adelph. 602. In the following the two substantives are different terms for the same person or thing: Et : Phorm. 35; And. 571, 813. Atque: Hec. 334. Cato. Que: de Agr. cult. p. 11, 10; 13, 18; 32, 8 ; 58, 9 ; 87. 16 ; 95. 9 ; 99. I ; Fragm. p. 87, 13. Et : de Agr. cult. p. 43, 13; 45. 9; 46.8; 79, I, 10; 83, 8; 90, 18; loi, 11; Fragm. p. 48. 15 ; 49, 2. Atque : Fragm. p. 22, i ; 23, 6 ; 33, 5 ; 41, 7, 8 ; 54. I ; 65,6; 74, 3; 83, 7. 20. In this section are included instances of two substantives united, of which one is a special and the other a general term. In the inscriptions the common particle used for adding the general to the special is que; in Cato the usual particle is et ; in Terence but few instances occur. To add the special to the general, both the inscriptions and Cato commonly have que, while Terence never uses que, but nearly always et. In view of this fact, the following statement from Reisig (Lat. Sprachwissenschaft, §233, p. 197) becomes incomprehensible : Da nun das que iiberall einen vermehrenden Sinn in sich schliesst, so folgt daraus, dass in der Kopulation mit que nicht Dinge verbunden werden konnen, wovon das zweite schon in dem ersten enthalten ist. («) General added to special. Inscriptions. Que : C. 34, magna sapientia multasque virtutes ; 195; 198,50,79; 200,93; 206, 69; 1008, 15, ameiceis noteisque ; 1418. Et . 1155; 1253. Atque : 1480. Terence, ig^i? •' Haut. 386. ^/.- Adelph. 19. ^/$^?i!f.- Adelph. 89 ; Hec. 396. Cato. Que : de Agr. cult. p. 40, 16. Et : p. 28, 14 ; 42, 2 ; 86, 4 ; 93. 5- Atque : Fragm. p. 41, 9 ; 55, 9. (J>) Special added to general. Inscriptions. Que : C. 195, 77, aera stipendiaque ; 200,91, bona agrumque; 1425. Neither et nor atque is thus used unless C. loio may be so considered : in dies et horas. Terence. Que not used. Et: Haut. 112; Phorm. 1049; Adelph. 971; And. 119, 558. Atque: Eun. 238; Phorm. 34. Ac : Adelph. 442. Cato. Que: de Agr. cult. p. 17, 24; 18, 14, pabulum lupi- numque; 82, 16, 18; 97,6; 100, i; loi, i. Et : p. 18,3,4; 58, I. Atque not used. 32 Que, Et, Atque in the Inscriptions 21. Substantives are sometimes united, of which one is, in meaning, directly opposed to the other. Both extremes are thus included. Cato uses only et ; inscriptions only atque (only once) ; Terence uses all three, but chiefly et and atque. Inscriptions. Only atque : C. 1008, 7, commoda atque incom- moda. Terence. Que: Eun. 193, dies noctesque. jEt: Eun. 1079; Adelph. 957, animo et corpore. Atque, regularly with deus and homo : Haut. 61 ; Hec. 198 ; And. 246 ; Phorm. 764. In the fol- lowing the substantives are not necessarily opposed in meaning, but are merely contrasted by the author: Phorm. 34, actoris virtus — bonitasque vestra ; Hec. 818. Et : Haut. 189; Phorm. 199; Adelph. 340; And. 540; Pronouns contrasted : Phorm. 442, me et se ; Hec. 338 ; Adelph. 340, 558, 566 ; And. 868 ; Phorm. 167. Atque, only with contrasted pronouns : Haut. 653, te atque illam ; Haut. 1030 ; Phorm. 368 ; Hec. 384 ; Adelph. 493 ; And. 233. 689. Cato. Only et used : de Agr. cult. p. 51, 9 ; 73, i ; 94, 13. 22. The uses of the copulative conjunctions with adjectives may conveniently be considered under a single section. We are met at the outset by the singular fact that, in the whole body of inscriptions, only three instances are found of que, et, or atque connecting two adjectives, though in two other cases one word is an adjective. In Terence and Cato, on the other hand, adjectives thus connected are very common, and in both writers the connective most frequently used is et. The two authors differ, however, in this respect, viz. Terence is particularly fond of joining adjectives of cognate meaning, while Cato is fondest of combining adjectives of contrasted, or at least wholly different, meaning. Inscriptions. Que : C. 200, 29, Latino peregrinoque ; 199, 66; privatus vectigalisque ; 200, 49, privatus vectigalisque ; one word an adjective : C. 205, XXI 20, id ius ratumque esto. Ac : looS, 14, incunda ac voluptatei fuel. (fl) Adjectives of cognate meaning. Terence. Qrie : Eun. 136, 419, perditum miserumque ; Eun. 487 ; Phorm. 164 ; Hec. 761, 848. Et: And. 36, 132, 229, 619, 956; Eun. 318, 682, loii ; Haut. 297, 327, 521, 580, 707, 797; Hec. 161, 472 (twice), 841; Adelph. 930, 986. Atque: And. 137, 274 ; Eun. 643, scelerosum atque inpium; Haut. 122,123, 633! Phorm. 239, 339, 499 ; Hec. 457; Adelph. 375. Ac: And. 123; Hec. 152; Adelph. 95, 756. of the Republic, in Terence, and in Cato. 33 Cato. Que: Fragm. p. 82, 3, celeris properaque. Et : de Agr. cult. p. II, 12, periculosum et calamitosum ; 18, 16; Fragm. p. 21, 6 ; 54, 2. Atque: Fragm. p. 14, 7; 19, 13 ; 38, 9; 39, 3. (J>) Of contrasted meaning. Terence. Que : Phorm. 376, te indignas seque dignas; Adelph. 73, praesens absensque. Ei not used. Atque: Hec. 380, 769. Not contrasted, but of wholly different meaning. Que : Phorm. 957, animo virili praesentique ; Eun. 73. £t: Haut. 120, 609. Atque: And. 811; Eun. 709; Phorm. 324, fortis atque amicus. Cato. Que : de Agr. cult. p. 13, 10, facta infectaque ; p. 44, i ; 87, 12 ; 94, I ; 107, 8. Et : p. 20, 8 ; 27, 6, maioris et minoris ; 29, 13.16; 38,1; 81,15; 93.10; 104,18; Fragm. p. 85, 4. Atque not used except with numerals: Fragm. p. 11, i, terna et quaterna ; 36, 5. Of wholly different meaning. Que : de Agr. cult. p. 11, 14; 15, 18; 41, 17; Fragm. p. 85, 6. Et : de Agr. cult. p. 18, 18 ; 19, 4 ; 20, 8, 12 ; 21, 5, 8, 9 ; 23, 14 ; 27, 17 ; 42, 9 ; 44, 20 ; 46, 9; 50, 6; 53, 14, 15; 54, 6; 78, 9 ; 85, 10; 103, i. Atque: de Agr. cult. p. 51, i ; Fragm. p. 28, 4 ; 42, 3. if) General added to special. Terence. Que: Haut. 788, aequi bonique ; Phorm. 637. Et : Haut. 649; Phorm. 451, 1008; Adelph. 987. Atque (ac) : Haut. 839, iniusta ac prava. Cato. Only et : de Agr. cult. p. 15, 14, viridius et melius ; 15, 16; 71, 2. {d) Special added to general. Terence. Que not used. Et : Eun. 926; Haut. 523; Phorm. 228, 767. Atque: Haut. 642, bonum atque aequom; Haut. 704; Phorm. 131, 497. Sometimes the second adjective is, rather, explanatory of the first. Et: Phorm. 623, erus liberalis et fugitans litium ; Adelph. 251. Atque: Eun. 938. Cato. Que: de Agr. cult. p. 12, 10; 15, 18. Et: p. 60, 14, bonam liquidam. Atque : Fragm. p. 39, 4, 7 ; 43, 3. Explana- tory : Que: de Agr. cult. p. 80, 15 ; 11, 11 ; 52,68. Et: 19,17; Fragm. p. 77, 3. (e) Adding cause or reason. Terence. jg>«e not used. Et : Haut. 438, leui et victo animo {leui because victo). Atque : Hec. 377, incredibili re atque atroci (because so atrocious). Cato. No example. (/) Adding result. Terence, jgas not used. ^/.- And. 953, magis ex sese et maius (more closely connected with himself and so of greater importance). Atque : Adelph. 849. Cato. Que : de Agr. cult. p. 37, 3, percoctum siccumque. Et : P- 29, 15 ; 42, 5 ; 72, 17- 34 Que, Ei, Atque in the Inscriptions i^g) Adding accessory notion. Terence. ^«« not used. Et: Haul. 704 ; Hec. 464. Atque : Hec. 457 ; Adelph. 403. Ac : And. 337, 591. Cato. Que not used. Et : de Agr. cult. p. 57, 9, 10. Atque: Fragm. p. 55, 10. (Ji) Adding adversative notion. Terence. Que not used. Et : Eun. 430, imprudenti et libero. Atque : Adelph. 155. Cato. Only one example : de Agr. cult. p. 12, 17. 23. As the preceding section was devoted to adjectives, so this is devoted to adverbs. And as in that place we noticed the absence of examples from inscriptions, so here we notice the same peculiarity in a still more marked degree. I have noticed in the whole body of inscriptions but a single instance of the union of two adverbs. This occurs in C. loii, 11, plus superaque. An adverb is joined to a substantive in two instances : C. 206, 3, item isdemque diebus ; 206, 5, item eademque omnia. In Terence, too, adverbs joined by et, que, or atque are of comparatively rare occurrence. In Cato adverbs thus joined are of still less frequent occurrence. (a) Of cognate meaning. Terence. Que: Haut. 1046, nimis graviter nimisque inhumane ; Adelph. 663, duriter inmisericordi- terque ; Haut. 594; Eun. 507. Et : Phorm. 344, 1047; Haut. 870. Atque: Eun. 56. Ac: And. 74; Haut. 957; Adelph. 45. Cato. Only que : de Agr. cult. p. 58, 3 ; Fragm. 68, 4. (J)) Of contrasted meaning. Terence. Only atque : Eun. 105, hac atque iliac (hither and thither) ; Haut. 578. Cato. Only que: de Agr. cult. p. 40, 19; Fragm. 23, 17. Not contrasted, but wholly different. Et : de Agr. cult. p. 56, 6 ; 73, 18. Atque: Fragm. p. 69, 4. {c) Special added to general. Terence. Only^^.- Haut. 226, bene et pudice; And. 274; Eun. 416; Hec. 857; Adelph. 953. Cato. Only once : de Agr. cult. p. 62, 4, bene et otiose. {cT) Adding a notion that is logically subordinate. Only once : Haut. 58, ut te audacter moneam et familiariter (as a friend would do). Not in Cato. {e) Adding accessory notion. Terence. Que not used. Et : Haut. 1 14, saepe eadem et graviter audiendo ; Hec. 240. Ac : Eun. 175, 915; Haut. 344; Hec. 552. Cato. Qtie : de Agr. cult. p. 89, 12, Et : p. 16, 7; 56, 3; Fragm. p. 9, 13. of the Republic, in Terence, and in Cato. 35 24. " Hendiadys," says Gildersleeve (Lat. Gram. 695), " consists in giving an analysis instead of a complex, in putting two sub- stantives connected by a copulative conjunction, instead of one substantive and an adjective, or attributive genitive." I find a few instances of this, mostly in Terence and Cato. Inscriptions. Que not used. E( : C. 1050, 5, ob fidelitate et oficeis (faithful services). Atgue : C. 33, gloria atque ingenium. TeRENCE. Que: Hec. 226, 406, vita solitudoque (lonely life). Et: And. 560; Eun. 1039, 1090; Phorm. 5; Adelph. 263, 837. Atque: Eun. 327; And. 843; Hec. 23, 225; Adelph. 42. Ac. And. 12, 75. The following, though not so clearly answering the above definition of hendiadys, still have one notion logically sub- ordinate to the other : Que : Eun. 383, nos nostramque adules- centiam habent despicatam (despise us on account of our youth). St: Haut. 626, meministin me esse gravidam et mihi te edicere (when I was pregnant, etc.) ; And. 71 ; Eun. 425 ; Phorm. 7, 212, 500. 593. 943; Hec. 123; Adelph. 280, 675. Atque: Haut. 190; Hec. 123; Adelph. 280. Cato. Que: de Agr. cult. p. 13, 2; Fragm. p. 85, 3. Et : 28, 20; 72, 6; 79, 3; 91, I. Atque: Fragm. p. 19, 11 ; 42, 6; 43, 6; 58, 9. Less evident examples: Que: de Agr. cult. p. 37, 10; 39.7; 62, 16; 63, 2. Et: p. 28, 20; 29, 4; 93, 3; Fragm. p. 64, 5, famosus et suspitiosus (notoriously suspected). 25. In the preceding sections we have considered the con- junctions with respect to the character of the notions joined. It may not be uninteresting here to examine these passages again with a view solely to the number of members which these con- junctions connect and the combinations in which they occur. (a) Three members. Inscriptions. When a series consists of three words, the first two are most frequently left without a connective and the third is attached by que.^ I find only 5 instances (out of 27 in all) where the first two also have a connective." xxquexque : C. 195, 12; 204, II 15, 20; 603,5. s. et -x. x que : 206,5. laetxetx: C. 623 ; 1147,3; 1180; 1253, 4; 1260; 1429 ; 1432. Next in point of frequency are cases where the series con- sists of three clauses. Here exactly the opposite holds true, that each couplet has its connective. The first couplet is here without a connective only 4 times out of 25 : -- que -que most common: ' See §26 (3). ^ Single words are represented thus : x x ; whole clauses, thus : — . 36 Que, Et, Atque in the Inscriptions C. 197, 10; 198, 15, 18, 52; 200, 8, 98; 203, 25;, 204, I, 7, 20; 206, 15 ; 577, II 5 ; 202, 28; 206, 15, 79; 1268; 205, II 47. -et --que: 198,68; 201, 10; 577,11 16. -et-et-: 206,10; 199, 3. — que et- : 197, 21. - et — (no connective with last two). Terence, ^a^quexque : Only Eun. 801. x^^x^/x: Haut. 845. 875. 941; Phorm. 1047; Hec. 210, 472, 592, 837. :x.y.que atque x : Adelph. 603. x atque xetx: Haut. 704. x atque x atque x : Hec. 457. xacx atque x : Haut. 855. - et — que : Hec. 748. -et- et- : Haut. 244 ; Hec. 599 ; Adelph. 285, 994. - atque - et- : Adelph. 590. -atque -atque-: Phorm. 322. Cato. xxquexque : de Agr. cult. p. 11, 14; 15, 18. — que -que: p. 11, 13; 44, 5; 49, 14; 52, 15; 63, 2; 89, 2; 92, 16; 96,9. — queet-: p. 38, 14 ; 51,19; 53.1°; 109,4. xetxetx: p. 20, 8; 22, i; 42, 13; 47, 13; 63, 15; 74, II, 13; 75. 11; 79.12; 81, i; 90, 2; 99, 13, 17; 109,5; Fragm. p. 68, 10. -et-et-: de Agr. cult. p. 17, 19, 21 ; 19, 5, 13; 38, 4 ; 40, 8; 44, 7 ; 56, 7 ; 66, 10; 68,5; 69,12; 70,5, 14; 72,8, 11; 74,10; 75, 15; 76.12; 79,8,14; loi, 20; 108, 5, 7. -et — que: p. 38, 16; 45,15; 60,5; 61, 10; 65,14; 93,3. X atque X atque x: Fragm. p. 21, 8 ; 36,12; 42, 9 ; 55. 10. {U) Four members. Inscriptions. -- que -que- que : C. 205, 15; 206, 10; 202, I, 36. XX atque XX que i^vst two without con- nective) : C. 33. Terence. No example. Cato. xetxetxetx: de Agr. cult. p. 42, 3; 50, 5 ; 60, 12; 68, 18; 69, 2; 73, 2; 103, 6. -et- et-et-: p. 29, 6; 75, 3. — que - que - que : p. 50, 12. - atque - atque - atque - : Fragm . P- 33. I- {c) Five members. Inscriptions. — que atque — que atque- : C. 196, 23. — que\et'\ — que-que: 198,77. -et-et-et-et-: 199.38. xetxetxetxetx: 199,39; 569. xetxetxetxx que : 1059. 3- X etxetxxque etx: 1065, 4. Terence. No example. Cato. xetxetxetxetx: de Agr. cult. p. 100,4. -et-et-et-et-: p. 102, 3. ((/) Six or more members. Inscriptions, --que -que -que - que - que : C. 202, 10. — que - que - que - que - que - que : C. 195. Terence. No example. Cato. — que - - que - que - que (connective omitted between 2d and 3d member): Fragm. p. 19, 10. -et-et- et-et -et-et-: de Agr. cult. p. loi, 7. of the Republic, in Terence, and in Cato. 37 In classical Latin this heaping up of qne, as shown under (4), was unknown (Drager, Hist. Synt. II, p. 37, 5), et being the regular connective in polysyndeta. It will be observed that Cato is particularly fond of polysyndeton, as Terence is particularly averse to it. For other peculiarities see §26. 26. Sometimes a connective is used only with the last term of a series. In the inscriptions this connective is almost invariably que. (a) Series of three. Inscriptions, xxxque: C. 33 ; 200,86; 46; 206, no; 589; 541 ; 198, 4, 6, 14 (twice); 200, 85, 100; 203, 23, 27; 206, 145; 551 ; 577, III I ; 1270; 1555. que: 198, 44 ; 199. 44 ; 200, 82 ; 205, XXI 7 ; XXII 46 ; 542, 5 ; 577, II i ; 603, 8. xxaiquex: perhaps 1008, 15, vobeis, viro atque ameiceis. Terence, xxxque: Hec. 92. xxeix: And. 24; Eun. 921 ; Adelph. 988; xxaiquex: Haut. 526 ; 893. xxacx: Adelph, 846. --atque-: And. 594 ; Phorm. 309. Cato. xxxque: de Agr. cult. p. 13, 13; 81, 14; 87, 11, 17, 18; 88,4; 52,13; 77,4. xxi?^x: p. 46, 16; 64,7; 101,18. que: p. 41, 11, 13; 48, 5. 7 ; 53- 4 ; 59. 15; 62, 2; 63. 11; 87,4, 16; 88, 12; 95, i; 97, II, 15; Fragm. p. 107, 16. — et-: de Agr. cult. p. 36, 7 ; 51, 18 ; 56, 5 ; 62, 4 ; 66, 17 ; 71, 14 ; 75, 11; 76, 8 ; 106, 6. Cato is also fond of connecting each of the last two cotiplets of a series — a peculiarity which I have not noticed elsewhere, xxxquexque : de Agr. cult. p. 81, 15. xxetx-que : 43, 6. que et-: 50, 8. que -que: 51, 20; 65, i. — et-et-: 74,2; also, -