Je 3 ‘26R The CLASSICAL QUARTERLY and CLASSICAL REVIEW are the Organs of the Classical Association. The QUARTERLY is published in January, April, and October, the last issue being a double number; the REVIEW in February, May, July, September, November, and December. THE CLASSICAL QUARTERLY EDITED BY EDWARD V. ARNOLD, Litt.D., Bryn Seiriol, Bangor, N. Wales; F. W. HALL, M.A., St. John’s College, Oxford. ; Board of Management ; Prof. J. F. DOBSON, M.A., Chairman. Prof. A. C. CLARE, Litt.D., F.B.A., Hon. Treasurer. L. WHIBLEY, M.A. Hon. Secretary. With the Chairman, Trea- J. P. POSTGATE, Litt.D., P.B.A. Prof, R. 8. weak PBA suver, and Secretary, repre- Representing the Cambridge Philological Soctety. cs gl Og tae senting the Council of the J. BELL, M.A. Prof. H. WILLIAMSON, M.A. Classical Association. Represent ng the Oxford Philological Society. With the co-operation of Prof. Wm. GARDNER HALE, University of Chicago; Prof. MAURICE HuTToON, Toronto; Prof. T. G. TUCKER, Litt. D., Melbourne. Vol. XX APRIL, 1926 CONTENTS PAGE A. L. Pack 57 J. D. Massotr 72 ANAXAGORAS AND THE PARTS ARISTOTLE AND THE XQ2PISMO® or PLato . A. C. Pearson, WILtiAM RipGeway 8o, 81 -1l. L. Acar 82 H. C, Nuttinc 86 . QO. L. RicHmMonp 89 A. W. GoMME 97 W. W. Tarn’ 98 F. H. Corson V’. M. Linpsay A. D. Nock Bee Bee i a we ee THE (Homeric) Hyruy ro Hermes CaTviLtvs VI. 1 saga. THe Cywruia Two Notes on HERODOTVS PoLyBIvVs ANL A LITERARY COMMONPLACE ‘TENAX PRoOpOsSITI’ GLEANINGS FROM GLOSSARIES AND SCHOLIA . INTRARE SVB 1VGVM SUMMARIES OF PERIODICALS: LITERATURE AND GENERAL , LANGUAGE ek oO alee LONDON: JOHN MURRAY, ALBEMARLE STREET, W. New York: G. E. STECHERT & CO., 31-33, EAST torn STREET. Yearly Subscription 16/- post free ; U.S.A. $4. Combined Yearly Subscription for the CLASSICAL QUARTERLY and the CLASSICAL REVIEW, 25/- post free; U.S.A. $6.25. Price for single numbers. 4/- net, or 4/2 post free. THE CLASSICAL ASSOCIATION THE objects of the Classical Association are to promote the development and maintain the well-being of classical studies, and in particular (a) to impress upon public opinion the claim of such studies to an eminent place in the national scheme of education; (b) to improve the practice of classical teaching; (c) to encourage investigation and call attention to new discoveries; (d) to create opportunities for intercourse among lovers of classical learning. Membership of the Association is open to men and women alike. The annual subscription is §s. (life composition, £3 15s.), and there is an entrance fee of 5s. (not charged to Libraries). Members receive a copy of the annual Proceedings of the Association and, on payment of 2s. 6d., of The Year's Work in Classical Studies (both post free). They may also obtain the Classical Review and Classical Quarterly at reduced prices, provided that the subscriptions be paid before January 31st in each year. Subscriptions sent in later than that date must be at the rates offered to the general public. Inquiries and applications for membership should be addressed to the Hon. Treasurer, Prof. M. T. Smiley, University College, London; or to either of the Hon. Secretaries, Dr. E. Norman Gardiner, 186, Woodstock Road, Oxford, and Mr. C. T. Seltman, Queens’ College, Cambridge; or to the Hon. Secretary of any of the district branches—viz., Mr. S. K. Johnson, The University, Manchester; Dr. G. A. Auden, Education Office, Margaret Street, Birmingham; Miss F.M. Rees, The University, Liverpool; Professor Granger, Uni- versity College, Nottingham; Mr. R. W. Jackson, 16, Freemantle Square, Cotham, Bristol; Mr. Basil Anderton, The Public Library, New Bridge Street, Newcastle- upon-Tyne; Miss E. J. Vaughan, Westfield College, Hampstead; Mr. W. E. Nash- Williams, University College, Cardiff; Mr. E. E. Bibby, The University, Leeds; Miss E. L. Topham, Firth Park Secondary School, Sheffield; Mr. E. D. T. Jenkins, University College, Aberystwyth; Miss M. E. Lees, University College, Bangor; Miss B. A. Leary, County School for Girls, Ashford; Mr. S. Price, 33, Walton Well Road, Oxford; Mr. J].M. Gaussen, 16, Lewes Crescent, Brighton, Mr. T. A. Sinclair, University College, Southampton; Mr. F. Fletcher, University College, Exeter; and Mr. C. W. Baty, 49, Waterloo Road, Bedford. Che Society for the Promotion of Hellenic Studies. Annual Subscription . ONE GUINEA. Entrance Fee - ONE GUINEA. PRESIDENT—WALTER LEAF, Esgo., Litt.D., D-Litt. OBJECTS.—The Society for the Promotion of Hellenic Studies was founded in 1879 for the following objects :— I.—To advance th2study of Greek language, literature, and art, and to illustrate the history of the Greek race in the ancient, Byzantine, and Neo-Hellenic periods, by the publication of memoirs and unedited documents or monuments in a Journal to be issued periodically. II.—To collect drawings, facsimiles, transcripts, plans, and photographs of Greek inscriptions, MSS., works of art, ancient sites and remains, and with this view to invite travellers to communicate to the Society notes or sketches of archzological and topographical interest. To organize means by which members of the Society may have increased facilities for visiting ancient sites and pursuing archzological researches in countries which, at any time, have been ti:2 sites of Hellenic civilization. Application for membership, or for information about the Society, should be addressed to the Secretary, at 19 Bloomsbury Square, W.C. 1. The Journal of Hellenic Studies and other publications of the Society are issued by Messrs. Macmillan & Co,, Ltd., and can be purchased from any bookseller, the published price to non-members being 30s. net per volume or 15s. net each part. Members are entitled to receive a copy of the Journal on publication during their membership, and can purchase back volumes at the subscription rate of 21s. per annum. Members also have the use of a well-equipped library at 19 Bloomsbury Square and of a large collection of lantern-slides illustrative of Classical Studies, t and ; upon tional (c) to create innual of 5s. edings assical assical before must o the or to stock o the , The Street, , Uni- tham, astle- Nash- eeds; D. T. ollege, ce, 33; shton, tcher, ord. 4 5. INEA. for the 1istory ication ically. potions, lers to iphical risiting time, to the fessrs. 2m bers of the ription 1. large o ease we ex than and of dat by lat bewilc @ ( | And c which have <¢ A term ; must | in eve Portio A the la really are, a ‘com (frag shall a Every it neve THE CLASSICAL QUARTERLY APRIL, 1926. ANAXAGORAS AND THE PARTS. THE great number of contradictory statements which confront us when we examine the various explanations of Anaxagoras’ philosophy make it more than usually important to decide what is to be admitted as first-hand evidence and what is not. I purpose, then, to begin by accepting the barest minimum of data, and I shall try to exclude any direct comments upon Anaxagoras’ work by later writers. Sufficient justification for such a course may be found in the bewildering masses of confusion which have gathered around his teaching. Our data, then, will be: (1) The problem which Anaxagoras had before him—viz. To account for Becoming and Change in view of recent developments of philo- sophical theory ; (2) The actual fragments of Anaxagoras’ work ; and (3) The existence of the words opolopépera, opotouepH, as applied to his philosophical theory. And our task will be to arrive at an interpretation and explanation of these which will also explain how the different interpretations of his philosophy may have arisen. Argument.—My argument will be that Anaxagoras uses potpa as a technical term; that this explains the origin of the term oyosopépeca; and thus opotopépera must be explained by Anaxagoras’ own assertion, ‘ There is a potpa of everything in everything. Hence his doctrine is that all things have similar Parts or Portions—i.e. any one has the same Parts or Portions as any other. Anaxagoras’ Data.—In the first place, then, Anaxagoras begins by accepting the law of Parmenides. No thing comes into being or passes out of being: what veally happens ts that a thing commingles, comes together, out of things that already are, and separates, disperses, mto the same again. Strictly speaking, we should say ‘commingles’ for ‘comes into being,’ and ‘ disperses’ for ‘passes out of being’ (fragm. 17, Diels), That is, all things that now are have always been, and shall always be, what they are now. That is the force of ao éévrwv ypnudtov. Every thing has always existed as such." And however small a thing may get, it never ceases to exist. Nor ts there a least of what is small, but there 1s always a 1 Cf. pp. 66, 68. NO, II. VOL. XX. 58 A. L. PECK smaller, for that which is cannot but be (fr. 3). Similarly, a thing may grow to any size. Once there, always there. But there is also always something greater than what is great. Nothing ever begins or comes to an end; it merely gets bigger and smaller. But, he continues immediately, though its size varies, it always has the same number, the same amount. [The great] ts equal to the small in number (rrAO0s) : it is in comparison with ttself that each thing ts great or small (fr. 3). We naturally ask, The same amount, the same number, of what ? Anaxagoras gives us his answer in fr.6. He says: And since the Portions (or Parts) of the great and the Portions of the small are equal in number, for this reason, too, all things will be in everything. Nor can things exist asunder, but all things contain a Portion of everything. Because there cannot be a smallest, therefore things cannot be sundered, and no thing can be by ttself; all things are together even now, just as they were at the beginning. And there are many things in all things: there is an equal number (mrjO0s) of the things which separate off in great things and in small things (fr. 6). Again he says: And there ave many Paris of many things. But no thing ts altogether separated off nor distinguished from any other thing except Mind (fr. 12). Again: We must belteve that all and sundry are in every thing that has ‘ come together’ (=the popular ‘ come into being’), 1.¢. that there are in them Seeds of all things, having all sorts of shapes and colours and tastes (fr. 4). And a little further on he says : When all things were together, not even any colour was discernible, for the mixture of all things prevented it—the mixture of the mozst and the dry, the warm and the cold, the bright and the dark ; of a great deal of Earth that was in it; and an infinite number of Seeds all entirely different from one another (fr. 4). [In passing, cf. fr. 12, where he is saying that Mind set all things in order: this revolution in which now revolve the Stars and the Sun and the Moon, and the Air and the Ether that separate off.... And the dense gets separated off from the rare, the warm from the cold, the bright from the dark, and the dry from the moist. And there are many Parts of many things (fr. 12). Here the same three divisions appear: (1) The ‘Opposites. (2) The ‘ Elements.’ (3) The ‘Seeds’ or ‘Parts.’] And as it was in the primal mixture, so it is now (see above): There ave many things in all things; there is an equal number of the things which separate off in great things and in small things altke (fr. 6). We cannot see all these Parts with our eyes (Aetius I. 3; Diels A 46), for our senses are weak, and we must not rely upon them for the truth (fr. 21), although as far as they go they are invaluable (fr. 21a). The Original Mixture.—We can now try to reconstruct Anaxagoras’ theory of the universe. To begin with, all things—notice that he assumes them to be already in existence—were so small that nothing was visible (fr. 1). Every- thing was depov; that is, it was as yet nothing in particular, there was 1 Hence a thing is always there, not merely else. E.g. salt will always be salt ; you cannot in the sense that the materials out of whichit divide itso far as to get a piece so small that the will be or was composed are always there, but next division will give N and Cl, neither of which it is always there as such: the pieces into which is salt. (I donot mean to imply that Anaxagoras you divide it will still be it and not something would actually have affirmed this of salt.) no ord and Et in bulk even w: were as only ‘S No Ether v large. things’ less dire which I that the (I) (2) (3) Taking (1) (2) (3) Cosi Anaxagc fragmen of the « Anaxago of whos that he physiolo, chiefly c tion to p The cosmolo; nothing instance, where* th further p I take it of these a 1 This, I Earth, whic is now ': cf, Hippol. Diels A 42; A. L. PECK 60 From Mists (vedeXov, which are a sort of Air) Water is separated off, and from Water Earth. From Earth Stones are formed by solidification by reason of the cold. Thus it was out of these ‘Elements’ that the heavenly bodies were formed. The Sun, the Moon, and the Stars were fiery stones (see Hippol. vefut. 1. 8.6; Diels A 42), carried round in the revolution of the Ether. The Sun and the Moon turned back when they had gone a certain distance from the celestial Equator owing to the force of the Air (¢bid.9). Andsoon. All this is in the old Milesian vein ; there is nothing here which Anaxagoras would claim as original: he was not particularly interested in such matters. Notice above all that there is no mention here of the peculiar doctrine of the OMoLopepn. Physiology.—We shall return to this point later. For the present, we will pass on to the physiological part of Anaxagoras’ philosophy, for this difference between the heavenly bodies and the ‘ Elements’ on the one hand and the rest of creation on the other seems to be due to ‘the distinction which Anaxagoras laid down between animate and inanimate things.’ Mind, it appears, is con- cerned pre-eminently with those things that have life. True, Anaxagoras says that Mind is where everything else is, even now, in the surrounding mass, and in the things which have been combined together and those which have been separated off (fr. 14);1 but he also says: There 1s a Part of everything in everything except Mind ; and there are things in which there is Mind too (fr. 11). This seems to be more from his heart. After enlarging upon the former point, about the Parts, he continues (fr. 12): Mind has all knowledge about everything and is strongest of all. But as for the things that have life, be they small or great, Mind controls them all, Thus, although Mind controlled and ordered the original mixture and the revolution and the Sun, Moon, Stars, Air, and Ether, its main function is to control the things that have life, and their combinations and dissolutions. The Process of Becoming as Explained Theoretically.—Hence, the contents of the original mixture which concern us at the moment are the ‘ Opposites’ and the ‘Seeds’: they are what is left when the ‘ Elements ’"—which always remain amevpa—are excepted. At this point Mind causes the whole mixture to revolve (7reptx@pnots); and when Mind causes the zepsywpnois there takes place a Sidxprous—a Sudxpiots which is also a cvpywkis. It is not a dedepiois merely (so that, e.g., all the gold particles come together to form gold); for no complete dudxpicts can ever take place, in the sense that anything will be completely on its own, apart from everything else; for whatever a thing is, it must contain in itself a Part of every other thing. The dsdxpiots consists in the things being got into order out of confusion, and in their becoming distinct; the cvppuéss consists in the congregating of Parts of all things in everything, so that things may still be wdvra ood (fr. 6), and in the building up of things. that he ‘dropped’ Mind after the initial move- ment of the world; although, of course, this is quite clearly inconsistent with Anaxagoras’ own words in frs, 12, 14 (tr. on p. 62). Cf. Arist. Met, A 984b 15 (tr. p. 61). 1 Curiously enough, although Anaxagoras made his name by ‘setting Mind over’ the physical world at large, he was chiefly interested in organic physics, and therefore there is probably some justification for Plato’s complaint takes He sa (fx. X32; maint: he ma: Aristot who si cause ¢ the w (Metap (a organi animal means collect proces: mainte be see! intellig than n think, ‘nothi of Min in that (6 actuall nouris! constit cold, fe becam TEPLXG: from t other ; from tl Thus 1 would Oppos signific off, and from veason of the bodies were (see Hippol. Ether. The istance from dsoon. All igoras would ers. Notice trine of the sent, we will is difference and the rest Anaxagoras ears, iS Con- merely (so 10 complete mpletely on - contain in lings being 1e cvppéts that things initial move- course, this is aXxagoras’ own ). Cf. Arist, ANAXAGORAS AND THE PARTS 61 This brings us to the next point. How are things built up? Anaxagoras takes his cue from the organic world of animals and other living creatures. He saw that Mind controls all things that have life, both great and small (fr. 12); and he extended this idea in two directions: (1) As we have seen, he maintained that Mind controlled the Universe as a whole; (2) as we shall see, he maintained that Mind controlled the subordinate processes of animals too. Mind in the ff Mind 5 Mind in the Universe as é€ 4 mn ? > subordinate processes a whole. \ animals. ¥ of animals. Aristotle bears witness on this point. He says: ‘So when a man arose who said that Mind was present im Nature, just as in animals, and was the cause of its order and all its arrangement, that man stood out like one speaking the words of soberness after those who before him had idly babbled’ (Metaph. A 984b 15). (a) The Seeds—In the same way Anaxagoras took another hint from the organic world to explain the apparent process of Becoming. The Mind in an animal or plant produces a fresh thing like the organism which it controls by means of seed. Nowa thing grows out of a seed not by a mere process of collection or accumulation of more seeds similar to the original one, but by a process of natural and organic development. Thus Anaxagoras does not maintain that ‘like joins to like’! to form a mass that will be large enough to be seen: his terminology suggests something quite different and much more intelligent. The word ‘Seed’ suggests a principle of organic growth rather than mere mechanical aggregation, just as the notion of Mind does. This, I think, is the clue to such phrases as ‘the Seeds of all things’ in fr. 4, and ‘nothing was visible because it was so small’ in fr. 1. The effect of the action of Mind upon the mixture was therefore to give the Seeds power to grow, and in that way to get large enough to be seen. (b) The Opposites.—But something else must be present before they actually can grow: for no seed grows until it gets nourishment. What is this nourishment which the Seeds get? I think it must clearly be the remaining constituent of the original mixture, viz. the ‘Opposites.’ The hot and the cold, for instance, were there; but like everything else they were ame:pa; they became distinct only when Mind produced the dsdxpioits by means of the mepixyopnots. But they, like the rest, did not suffer complete separation one from the other (fr. 12). The things in the one world are not sundered from each other ; no more ts either the warm cut off as with a hatchet from the cold or the cold from the warm (fr. 8). Wherever the cold was, there also the warm must be. Thus when the Seeds originally drew to themselves material for growth, they would draw the cold as well as the warm and so on throughout the list of Opposites. But the material itself was always there (though it had no significance: nothing could be distinguished in it as hot, as cold); and the 1 See below, p. 67 and following. 62 A.'L. PECK Seeds, too, were always there (though they, too, were insignificant: they had not yet begun to grow). Thus our theory explains Becoming. Change.—But once formed, once constituted in this way, things do not remain as they are. After they have reached their fulfilment they dissolve and change. (Notice here how appropriate our terminology is. A seed develops into a plant or an animal and finishes by producing seed again. A poipa is something that has to be fulfilled, to be developed: its nature is settled once for all, and cannot alter; yet it unfolds gradually.) We have thus the further problem of Change, and to explain this we must develop our doctrine of Seeds. If we find one thing turning into another, as, for instance, Bread turning into Flesh, that means that there must have been in the Bread a Seed of Flesh, because no thing can come into being; it must always have been there. We must assume, therefore, that there are in Bread Seeds of all the things into which Bread can possibly turn—that is, that there are Seeds of all things in it: there is a Part of everything in everything. Let us examine the matter more closely. The Parts of all things which are present in any piece of Bread were at the beginning in the original mixture; but they were scattered about; there was no order among them. Mind, however, which brought order into the mixture, and decreed all the combinations and dissolutions that ever have been and ever shall be (see fr. 12), decreed that this Seed or Part of Bread should develop and actually produce the Bread which it was capable of producing. But since nothing can exist all by itself, there were of necessity with the Part of Bread Parts of all other things, though of course the Part of Bread, and not they, was in power (=xparteiv). So Bread ‘came into being,’ by cvppeéis. But when Bread gets into the body, through the decree of the same Mind, which 1s even now where all other things are, and in them (fr. 14), controlling them (fr. 12), it gets transformed into Blood, Flesh, or Hair, as the case may be; and it can do this (1) because there were in it Parts of Flesh, Blood, Hair, and the rest, as well as a Part of Bread; and (2) because Mind has decreed that here, in the body, at this particular place, such Parts as Flesh, Blood, and Hair shall be in control. Thus the Part Hair now controls—almost ‘informs’ —the material which the Part Bread controlled before. So we have solved the puzzle of Change, how Hair can come out of what is apparently not Hair, and Flesh out of what is apparently not Flesh (fr. 10). The Doctrine of Similar Parts.—And because, as we have seen, if Change is to take place, there must be in each thing a Seed or Portion of every other thing as well into which it might possibly change, it follows that the Seeds or Parts in any thing will be exactly similar to the Seeds or Parts in any other thing; all things will have similar Parts. This doctrine is therefore the doctrine of Similar Parts (owovopépeva). The Process of Becoming as it Took Place in Actual Fact.—Now, as we have seen, the doctrine of Similar Parts seems only to apply to animate objects. But the two sides of philosophy cannot be left unconnected, and Anaxagoras made the connexion in the following way. Anaximander had held that animals (Diels | Now th pT animal: i4, 2: descen¢ Plants, to Ana: instanc from bs these [ (De ca ' Seeds. Pl status is to | Socrat as ‘ Ail thinki: Earth, that \ not w @UNY, pnoba DOTTED ‘ drop: (] Not < sort. and ¢ t: they had ings do not dissolve and red develops A poipa is settled once the further ne of Seeds. turning into ed of Flesh, there. We things into things in it: matter more Bread were ibout ; there der into the sr have been read should ' producing. th the Part ad, and not y cvpuprtis. same Mind, rolling them se may be; l, Hair, and screed that Blood, and t ‘informs’ ave solved y not Hair, Change is every other e Seeds or any other refore the ‘Ss we have te objects. Naxagoras held that ANAXAGORAS AND THE PARTS 63 animals first arose in the moist element. This Anaxagoras accepted (Diels A 113), and extended the principle to cover plants as well (A 117). Now things come out of Seeds. But whence came the Seeds ? Out of the Air. ‘Howbeit Anaxagoras, which also was surnamed the Atheist, taught that animals arose from Seeds which fell down from Heaven to earth’ (Ireneus II. 14, 2: Diels A 113). ‘Anaxagoras said that Air held the Seeds, which, descending with the Water, brought forth plants’ nee -cametand History of Plants, III. 1,4: Diels A117). Aristotle confirms this. He says: ae to Anaxagoras, the Elementals (crovyeta) were the Homoeomeries—I mean, for instance, Flesh, Bone, and each of such things—while Air and Fire, so far from being Elementals as Empedocles per contra maintained, were mixtures - these [i.e. such Elementals as Flesh and Bone] and of all the other mene (De caelo III. 302a). N.B. in passing: he equates Homoeomeries wit groom on Anaxagoras.—An unmistakable confirmation of the difference of status of Anaxagorean cosmology and Anaxagorean physiology and biology is to be found in Plato’s discussion in the Phaedo. It 1s plain that when Socrates says (98) that Anaxagoras brought in other causes beside ners such as ‘Airs, Ethers, and Waters, and a great deal of similar nonsense,’ he 1s thinking of cosmology, for he has just before been talking of the state of the Earth, the Moon, and the Stars. He says he had expected Anaxagoras to say that Mind arranged them xata ro BéXticTov; whereas he ward says how, not why, Mind arranged them, viz. through Airs, etc. ov yap av more auTov @unv, PacKkovta ye UTO vod avta (the Earth and te Reawanty orange ren pnoOat, dAXnv tia adtois aitiav éreveyxKetv f Ott TO BEXTLGTOV aUTA OUTMS EXEL aomep exer (Phaedo 98A; cf. 97c). According to Socrates, then, Anaxagoras ‘drops’ Mind in two senses : (1) He makes no use of Mind; but first says it is the cause of all things, and then alleges all kinds of other causes. [This is probably a partially justified misinterpretation on the part of Socrates (see above, p. 5, note), which Aristotle aggravates, Met. A 4. 985 b 18. For see fr. 14: Mind .. . is where everything else is, etc.; and fr. 12: Mind set in order all things that were to be and all things that were and ave not now, and all things that are . . . even this revolution in which now revolve the Stars and the Sun and the Moon, etc. Mind has all knowledge about everything (Ibid.).] (2) A typically Platonic (Socratic) objection. Anaxagoras is not interested in morals, but in physics. He begins by raising our hopes with the mention of Mind, and then ascribes to Mind only physical powers, neglecting the ethical. Not a word is breathed about the doctrine of Homoeomeria or anything of the sort. That is because the discussion at the moment is limited to cosmology and astronomy. 64 A. L. PECK The Testimony of Aristotle—Let us now pass on to consider the chief statements of Aristotle on the physiological theory. They are these: (1) Met. A 984a: ‘ Anaxagoras says that the dpyai are unlimited. For he says that almost all the things which have similar parts, just like Water or Fire,’ come into being and pass out of being in this way, viz. by combination and dissolution alone ; absolutely, they neither come into being nor pass out of being, but persist eternal.’ (2) De caelo III. 302a: ‘ Anaxagoras goes contrary to Empedocles on the subject of the Elementals. Now Empedocles says that Fire, Earth, Air, and Water are the Elementals of bodies, and that everything is composed of them ; but Anaxagoras says the opposite, viz. that the things which have similar parts are the Elementals—I mean such things as Flesh, Bone, and the like—while Air and Fire are mixtures of these and of all other seeds, since each of them is a concourse of all things which have similar parts, though they cannot be seen. Wherefore all things come into being out of them. N.B.—By Ether Anaxagoras means the same thing as Fire.’ (3) Phys. III. 203a: ‘ Those who make the Elementals unlimited [sc. in number| affirm that the unlimited is continuous and without intervals, Anaxagoras making it consist of things which have similar parts, Democritus of the panspermia of shapes (cynudtov = eida@v = atoms). And Anaxagoras says that any one of the parts (uwopiwyv) is a mixture just like the whole (see fr. 6), since he noticed that anything comes into being out of anything.’ (4) De gen. e¢ corr. 314a: ‘ Anaxagoras postulates as Elementals things which have similar parts, such as Bone, Flesh, Marrow, and every other thing a part of which has the same name’ (sc. as the whole). (5) Met. X 106g9b: ‘ All things come into being out of what is—or, rather, out of what potentially is, but is not actually. That is the same as Anaxagoras’ One (better call it that than “ All-things-together’’) and the ‘‘ mixtures” of Empedocles and Anaximander.’ ‘Opotopeph Ambiguous.—In four of these passages Aristotle speaks of the things which have similar parts (ra oyovoyeph). And from the way in which he uses the term it does not seem that Aristotle himself introduced it into the subject. He always says, for instance, that Anaxagoras postulated that the Homioeomeries were the Elementals, and then proceeds to explain what the Homoeomeries were. He does not enumerate them first as he might easily and perhaps more naturally have done, and as he actually does when speaking of the four elements of Empedocles in the second passage translated above (wip xai Ta cvoTotxya TovTos). Rather the word seems to have already held the field, and to have been inherited and explained by Aristotle. Now it is true that Anaxagoras was interested chiefly in organic physics, and therefore we may suppose that Flesh, Bone, Marrow, and the rest, which Aristotle takes to be meant by Homoeomeries, would be in the fore in 1 The meaning of this phrase is of course with the next quotation. disputed ; as it stands it is clearly inconsistent Anaxag mentio what ts whethe Marrov have t repeats everytl admit, concret Aristot doctrin explan: develo} ‘ things how t! Aristot Aristot other | examp! were tl else—p nothin; them, Anaxag I Aristot interpr becaus mean t and Ae throug have tl is the existen could | O{LOLOpe everytl evident Aristot all, we an ans’ 1 It is goras ust r the chief e: d. For he e Water or ombination or pass out cles on the h, Air, and -d of them; milar parts like—while of them is cannot be -By Ether ted [sc. in intervals, Yemocritus goras says (see fr. 6), als things her thing or, rather, laxagoras’ tures” of ks of the in which | into the that the uin what tht easily speaking ed above ady held physics, t, which fore in ANAXAGORAS AND THE PARTS 65 Anaxagoras’ philosophy. We have a fragment in which two of them are mentioned: How can Hair come into being out of what is not Hair, or Flesh out of what is not Flesh ? (fr. 10). But I think if we were asked to say quite fairly whether the general impression we get from the fragments is that Flesh, Bone, Marrow, etc., are expressly indicated as the Elementals of matter, we should have to say that they are not. For if there is one thing which Anaxagoras repeats ad nauseam it is the doctrine of wotpav—there is a potpa of everything in everything—no thing is completely separate from any thing else. We must admit, I think, that Flesh, Bone, Marrow, etc., are an attempt to explain concretely what the yotpa doctrine propounds theoretically, rather than that Aristotle has coined the word oposoyeph, or transferred it from his own doctrine, to describe what Anaxagoras enumerated in concrete terms. If the explanation of Anaxagoras’ theory which I have put forward is correct, the development of the term is extremely natural. And if opocouepy is translated ‘things which have similar parts,’ as I have translated it above, it is easy to see how the term could be applicable both to Anaxagoras’ doctrine and to Aristotle’s own doctrine—whence our present confusion. The oposopeph of Aristotle were (see No. 4 above) things of which every part was similar to every other part of the same thing, the same all through, Bone, Flesh, Blood, for example. We should call them homogeneous. The opocoweph of Anaxagoras’ were things of which all the Parts were similar to all the Parts of everything else—poipa tavtds év mavtl éveotw. The ‘similar parts’ in Aristotle are nothing more than similar pieces: they had nothing definitely organic about them, since Aristotle’s thought was not dominated in the same way that Anaxagoras’ was by by the notion of Mind in everything. I do not of course wish to claim that in all the statements either of Aristotle or of the others who comment upon Anaxagoras owotouepns should be interpreted to mean, ‘having Parts similar to the Parts of everything else,’ because it is obvious in very many cases that the writers did not intend it to mean that, although in some cases they very clearly did (see below, Simplicius and Aetius). All I wish to claim is that until Aristotle introduced confusion through having an omocopueps theory in his own system, the word could and did have that meaning. So far as its application to Anaxagoras is concerned, that is the only justification for the word’s existence. And so, taking its mere existence as a significant fact, I have tried to show why and how the word could reasonably have come into use. Obviously we must explain the word Ouovouepns by the statement of Anaxagoras himself that there is a portion of everything in everything, and not vice versa. In view of the conflict of evidence, we must not assume that the actual words of Anaxagoras and the Aristotelian interpretation of ouovowepys are even of equal weight. And above all, we must keep in mind the problem to which Anaxagoras was trying to find an answer. 1 It is not necessary to suppose that Anaxa- n_ se before Aristotle. goras used this term, but only that it was already 66 A. L. PECK Thus, since all things (that is, all sensible objects) will on our showing be opooueph, I think it makes for greater clearness if we speak of the doctrine of opo.opépeca, the doctrine of the Similarity of Parts, rather than of opotopepy. Bearing this in mind, however, we may speak of the oworoyepy! as Elementals —that is, little collections of all the Parts. (Notice that these differ from a mavotrepyia such as Air in that Air is ‘unlimited’ and no Seed or Part in it has any control.) The opinion of Aetius on the matter is, if not weighty, at least interesting. He says (I. 3, 5): ‘ Anaxagoras, the son of Hegesiboulus, a citizen of Clazomenae, declared that the Elementals of all that exists were the things with similar parts (oporouépevar). He found it extremely difficult to see how anything could come into being without having been in existence all the time (é« rod yz dvTos) [i.e. as such], or to perish and cease to exist any longer [i.e. as such] (@0eipecOax és TO yy dv). Example. We eat food that is simple and uniform (addy xal povoeroy), such as bread and water, and by this food hair, veins, arteries, flesh, sinews, bones, and the other parts are nourished. Since this is the case, we are forced to admit that all existing things must be in the food we eat, and that the increase of all things that do exhibit increase comes from what is already in existence as such (é« t@v dévtwv). This means that in the food we mentioned there are parts productive of blood (udpia aipatos yevyntixa), sinews, bones, and the rest. Which parts were visible to reason and not to sight. We must not make the senses a test of every theory, and try to prove to their satisfaction how bread and water produce hair, etc. On the contrary, these parts are visible to reason alone. Hence, because the parts (uopsa) in the food were similar to the things which they produced he called them opoiopéperat, and declared that they were the dpyai of the things that are. The oposopépevar were therefore matter (fA) and the efficient cause was Mind which set all things in order.’ There is also an important passage in the pseudo-Hippocratean De uictu, which will be quoted below. This work confesses to borrowing from previous writers. There is, of course, an obvious objection to such a unit as an opovopépera. Surely a collection of all the Seeds or Parts cannot be indivisible? But on this point Anaxagoras is explicit and insistent. The things in this world are not cut off one from the other with a hatchet; there is a Part of everything in everything. Therefore an opotouépera is not divisible. The whole doctrine falls to the ground unless we maintain this. In fact, the action of Mind was to turn separate and orderless Seeds into ordered opotoyéperar—it did away with the ‘unlimitedness’ of the primal mixture. Take an example. When Bread is digested, the Parts of Flesh do not get separated from the Parts of Bone, Hair, etc., but the piece of matter possessing the form of Bread which gets carried to the place where Bone is in control—that is, where matter always 1 Thus each duooepés will be an example in practice of the doctrine of duoropépera. takes th always | carried t develop that gro suppose we are n showing the very Flesh.’ Lik without cosmolo for hold: ment. — commen Anz Codepov éFeyopn after a v Mind cz complet on to ex things t apparen Simplic: TH ev a TOTE [el of cour: Anaxagt TOV ved: puvTae & Sin early an moment the twc extreme if not a (I) (2) No earlier t he desc TUKVOV howing be loctrine of OjL0Lo Ep}. tlementals fer from a Part in it nteresting. ., declared ilar parts ould come vTOS) | 1.e. sipecOar és ATAHVY Kal ries, flesh, » Case, we , and that is already nentioned vs, bones, We must tisfaction parts are ood were erat, and OLopé pera h set all De uictu, previous DLO epeta. But on vorld are ything in doctrine d was to vay with 2n Bread of Bone, ich gets r always l. ANAXAGORAS AND THE PARTS 67 takes the form of Bone—gets transformed; because the Part of Bone that has always been in it now gets control. Similarly, another piece of Bread, being carried to the place where Blood always is, gets transformed into Blood, by the development of the Part of Blood which it has always had. It is in this sense that growth may be described as To Spuoov mpos TO dpovov. If, however, we suppose this to mean that Flesh grows by the addition of more actual Flesh, we are met with the difficulty that Bread is not Flesh but Bread, and on any showing there must be more Bread in it than Flesh or anything else. Indeed, the very fact we have to explain is that Flesh does ‘come out of what is not Flesh.’ Like to Like.—But the primary application of the ‘like to like’ principle is without doubt to cosmology. We have noticed the different status of cosmology and physiology in Anaxagoras’ philosophy, and we have seen reason for holding that the doctrine of Similar Parts applies only to the latter depart- ment. Now this is borne out by the fragments of Anaxagoras and by the commentators. The pertinent passages are these: Anaxagoras (frs. 15, 16, Diels): 7d wév muxvov cal Svepov Kal yuypov Kai TO Cohepov évOade auvexwpnoer, ev0a viv, TO 5é apatov Kal To Oepyov Kai to Enpov éFeywpnoev eis TO Tpdcw Tod aiépos. This follows immediately upon fr. 12, after a word or two by Simplicius. In fr. 12, Anaxagoras has been saying that Mind caused these various things to separate, and yet nothing separated off completely. Apparently, from what Simplicius says after fr. 15, he then went on to explain this more fully, showing how the four ‘ Elements’ were the first things to be produced, and the most simply constituted, while other things, apparently organic, were made later, and were of more complex structure. Simplicius’ words are (following immediately upon the above quotation): «al Ta pev apxYoeon Kal dmAoVeTAaTA aToKpiverOas éyet, GAXa SE TOUTWY cUVOETwWTEPA Tore pev cuptnyvucbat A€yer WS cUVOETA ToTe 5é aTroKpiverOat ws THY yhv (this of course is where animate creatures are actually found). ovtws yap dyow: Anaxagoras (fr. 16): amd TovTwy aroKkpivopévav cupTHyvUTaL yh* ex pev yap TaV veheAov Udwp atroxplvera, éx Sé Tod HdaTos yi, ex S€ THs ys MOoe cuptipy- vuvTat UTO TOU WuxXpod. Simplicius is clear that there are two sorts of things: one class is formed early and is simple; the other class is formed later and is complex. At the moment, however, he is concerned with the first class, which is inorganic, and the two quotations which he gives from Anaxagoras refer to this. It is extremely probable that the two were fairly close to one another in the original, if not adjacent. Thus Anaxagoras says that— (1) The dense, the moist, the cold, the dark, came together hither ; (2) The rare, the warm, the dry, went away where the Ether is. Now we see from fr. 1 that Anaxagoras considered Air and Ether to be earlier than anything else. And we learn from Theophrastus (Diels A 70) that he described the Ether as being pavovy AerTov and Oepydv, and the Air as muxvoy tayv and wWuypov. I think, therefore, that in fr. 15 we should not 68 A. L. PECK insert <1) >, but take the words to refer to the Air (see above, p. 59, note). Fr. 16 will then describe how the Air gradually solidified into Cloud, Water, Earth, and then Stones. The outcome of this is that those fragments which seem to indicate the complete separation of certain ‘Opposites’ unmistakably refer to inanimate objects. Diogenes (Diels A 1). Ascrappy account. If this is a case of like coming to like, as it seems to be, it clearly refers to cosmology. Plutarch (Diels A 12). Another reference to cosmology. Olympiodorus (Diels A Io). Hippolytus (Diels A 42). Cosmology. Aristotle (Diels A 88). The two passages of Simplicius and Aetius, which are clearly parallel, should be considered together (Diels A 45 and 46) : Simplicius (A 45): eds uév tH Tovav- Aetius (A 46): éd0xes yap avtT@ arro- THY Evvocay [sc. éxdorny opotopéperav | pwratov elvat, mas éx TOD pn OVTOS oMolws TH Ow TavTa éxovoay évuTrdp- | Sivatat TL yiverOa 7) POcipecOae és TO Xovra] o Avatayopas 7\Oev, Hyovpevos | pt ov. penodev éx TOU un OvTOS yivecOa. .. . , TPODNS dé Tis avdTis mpocdhepouevns tpodyy yoov mpordhepoucba atdiv oloy apTov TOANG Kal avopota yivetat, | Kal povoerdn, aptov Kai bdwp, Kal éx TapKes ooTd préBes veipa Tpixes dvuxes | TavTHS Tpéhetac OpiE prev apTnpia Kal TTEpa O€ eb OTW TUXOL Kal Képata, | cape vedpa doTa Kal TA AoLTTa popla. avgerar dé TO Guotov TH opoiw. ‘ dua TavTa é€v TH TpopH wmédaBev TOUT@Y OUV ryLVOMEVOY OMOdOYNTEOY eal Kal ev TH VoaTL, eb TOUTw Tpé- | STL ev TH TpOdH TH Mpochepopevy TavTa porto ra dévdpa, Evrov wal provov Kai | éotl ta dvta, Kai ex THY dyT@V TaVTA KapTov. avferar. Kal év éxeivn €oti TH TpOPH popia aipatos yevyntixd, Kal vevpwv co ae : Kal OoTéwy Kai TOV GAXwY. ..- . 10 TavTa €v Taow Eheye pEepixOat, év TovTos eat AOy@ OewpnTa popia. Kal THY Yyeveow KaTa Exxpiow yiverOat. | ao Ooty Gyola Ta pepNn Elvar EV TH Tpohh Tois yevvwpévors opmotopepetas auTas éxddecve. . . . Etc. | Etc. The meaning of av&erar 5€ 7d Gyotov TH opoiw in Simplicius is that Anaxagoras took it as an axiom that one thing did not turn into another, but that Parmenides’ statement was true. Nothing comes into existence absolutely. All things, therefore, exist already as such. It really means the same as the corresponding phrase in Aetius, é« Trav dvytwv tavta avéerat. With these passages must be taken— Simplicius (Diels A 41):-dnolv év 1H Staxpices Tod ameipov Ta auyyevh pépecOat pos GAANAa, Kai Ste pev ev TO TravTl ypuads Hr, yiverOa ypuvcor, OTL dé yj, yv> omoiws Sé Kal Trav GAXov ExacTorv, ws Ov yLVoMéevwV GAN’ éevuUTTAaPXOVTOV . . aittov tov voov. He has said just before: mdavta Ta opotopeph), olov idwp wip 7 Xpvaov, ayévnta pev elvar Kai aPOapra, paiverOar Sé ywwopeva , MPOTEpoy . \ > 4 , \ an / kal atoddupeva ovyxpice Kal Siaxpice: povov, TdvTwy pev ev TaoW EevovTOD, e 4 \ A. . 2 ~ > . A / €xaoTou € KaTa TO ETLKPATOUY EV AVT@ vapaxTnp.fopevov. F1 cosmol neither he is n Tl in the it is tc where words e €K ovUTEe 7 vTrapyX EXOn, & acum avrnr TO Oe av pw’ TOV pL TAUTA TH TEP éveln pi ov yas eKacTe bh} of wh necess T philos (as th thing transf tion. the tr a thir that v Wate the n realit: was | Emp perisl poun D. 59, note). bud, Water, indicate the . inanimate ike coming ly parallel, AaUT®@ atro- } un OVTOS > \ eo0au €s TO 0a amd DP, KAL €K LY aptnpia a popla. LoONOyNTEOV evn TaVTa ‘WY TaVvTa vi rp0b ai vevpwv ITA popta. yat €V TH oLopepetas is is that ther, but »solutely. 1e as the ouyyevn UT OV, OTL LPpXOVT@Y wOLomepy, rylvopeva €vOVTOD, ANAXAGORAS AND THE PARTS 69 From the whole passage one gathers that Simplicius is referring to cosmology, though that might not be clear from these two excerpts. However, neither of the examples he chooses, gold and earth, is organic. And therefore he is no doubt thinking of frs. 15 and 16. The key to the two parallel passages on p. 68 is, I think, to be found in the theory I have already expounded. An interesting confirmation of it is to be found in the pseudo-Hippocratic treatise mepi dvaitns, chapter 6, where the author is speaking of the Parts in connexion with growth, and his words are: e , \ \ , A > A # A 4 / \ e A éxaoTn 5€ Wuyn wélw Kal eXdoow Exovoa TrepipoiTa TA popla TA EwUTIS, ¥ Gé ov > , 6 / n , \ de ” a ovTe mpoabéctos ovTE adaipécvos Seopévn THY péepewy*> Kata dé av—now Tov iTapxXovTwv Kai peiwow Seopeva ywpns, Exacta Siatpynoceta, és Hvtwa av EXOn, wai Séyerar Ta mpooTinrovta. ov yap Suvatat TO wy OmoTpOTrOY év TotowW / , tad aovppopotct Ywpiorow eupévery* TaVaTaL wey yap ayvopmova* auyyiwopueva Sé > / , \ A / / s A , A , GrAAnAOLaL yiwwOoKEer pos 6 mpocifer: Tpocife: yap TO cUphopov TO cuphopo, To 5¢ aovpdopov trodepet Kal payxetat Kai SiadrAdooe aT’ adXrAnrAwY. Sia TOVTO avOpwtrov ~ruyn év avOpwr@ avfetat, é€v GrAXrw 8 ovddevis Kai TOV ad\rav Cowv oe , . i id > aad b] 9 > Ta e \ , » , TOV peyadkwv woavTws Oca addOla aT’ avTav vo Pins aTroxpivetar.. . tavta 5€... Woda cai travtoia [sc. Ta pépea] tpéhetai te cal avEetar Siaity TH wep avOpwtros. avayKn 5é Ta péped EXEL TAaYTA TA EloLOVTA* OUTLVOS yap MN > / a > b A > Xn > / Ld n A > 4 ¥ > / Eveln potpa €& apy7s, OVK av avénGein, ovTE TOAAHS TpOdis Emiovans oUTE OALYNS, , A A ov yap exes TO Tpocave~avopevov. éxov dé Tavta avkerar ev yopn TH éEwvTOV &KacTov. ... See also other passages of Book I., where there are a number of hints of which I have made use in filling up gaps in Anaxagoras. Caution is very necessary, because the work is obviously, and indeed confessedly, a medley. The first attempt to answer the criticism of Parmenides upon the old philosophy was made by Empedocles. It was no longer possible to maintain (as the early philosophers had done) that the whole world consisted of one thing in various forms, nor (as Heracleitus had done) that things could be transformed one into the other. Empedocles made the least possible altera- tion. He seems (says Professor Burnet, E.G.P. 228) to have turned each of the traditional ‘ opposites ’"—hot, cold, wet, dry—which lay ready to hand, into a thing that was real in Parmenides’ sense of real. He thus got four things that were ultimate, and did not admit of transformation, Fire, Air, Earth, and Water. Of these all consisted. Only these had a reality of their own. All the numberless forms which are presented to our senses are not ultimate realities. Indeed, no Greek philosopher had ever said that they were. That was the position to which Greek philosophy was at last to come. For Empedocles these numberless forms never even came into being and never perished (fr. 8), because (in modern language) they were not chemical com- pounds but mere mixtures. Thus they could have no real being of their own. 70 A. L. PECK They were a mere mingling of the four Roots, which alone were ultimate. But ‘Anaxagoras regarded this as inadequate’ (E.G.P. 262). He probably saw that there was no point in limiting the number of the ultimate forms of reality to four, especially as the four forms chosen by Empedocles had little or no reference to living organisms, in which Anaxagoras was chiefly interested. If we must have a plurality of ultimate forms of reality, why not make every one of the numberless things around us ultimately real? Hair, flesh, bread, and so on (fr. 10 D.)? And that is what Anaxagoras actually did. He made everything ultimately real: everything is always there. Just like a plant or an animal; when it is not there as a plant, it is there as a seed. And since, so far as we can see, everything can turn into anything else (Aristotle, Phys. A 4. 187b 1), there must be a seed of everything that is ultimately real in each of the things we see. Or, more strictly, each sensible phenomenon must contain in itself seeds of all the possible sensible phenomena, for the form of each of those is an ultimate form of being. Of what Things are there Parts >—If Anaxagoras had been asked, ‘ Of what things are there Parts?’ I imagine he would have answered that, generally speaking, there are Parts of such things as have life or have at some time had life. Clearly there will be Parts of Hair, Flesh, Bone, Blood, etc., and of vegetable substances, such as Bread, Wood, and so forth. There will not be Parts of Fire, Air, Earth, or Water, because these are merely fortuitous mixtures of all sorts of Parts, in contrast with the orderly mixtures such as Flesh or Bread. They are not organic, not organized. In a word, there are Parts of the things with which Mind is the most intimately concerned (see fr. 12, p. 60). There is, of course, no Part of Mind itself (fr. 11, 12), for Mind does not enter into anything as an ingredient of the mixture. Nor is it at all necessary that there should be a Part of Mind in everything; for things do not turn into Mind as (e.g.) Bread turns into Flesh. Metals, being a form of Earth, would be subject to the same rules as Earth and the other ‘ Elements ’—that is, solidification and rarefaction (fr. 16). Change by this means is, of course, unknown in the organic world with its system of Mind and Parts. It seems, in fact, to be the power of Mind i the organic things (fr. 14, see p. 62) that makes it possible for change to take place. Every sensible object has, in itself, a nucleus of all the Parts ; one of which Parts controls the whole object and makes it what it is. This is in virtue of Mind. If another Part gains control (that is, if the object seems to turn into another object), this too is in virtue of Mind, which (fr. 12) has all knowledge about everything. .. . The things that have life . . . , Mind controls themall.... Mind has knowledge of all the combinations and separations and sunderings that take place (n.b., present). Mind set in order all things that were to be, and all that were and are not now, and all that are. Anaxagoras, we remember, had said that there were in the original mixture—that is, that the material of all existing things was: (1) (2) I o Seeds o that diff need mc there ws says tha things t the War all the ' that th will of | the Op needed without adding the doc Sui alone i: mixture (1) (2) CHR re ultimate. fe probably te forms of iad little or interested. nake every esh, bread, He made plant or an And since, (Aristotle, ely real in enon must 1e form of ‘ Of what generally time had >. and of ill not be fortuitous 5s such as there are rned (see for Mind $s it at all 7s do not rules as refaction lic world s (fr. 14, sensible trols the another ct), this _ Sees lowledge resent). ow, and original ANAXAGORAS AND THE PARTS 71 (1) The moist and the dry, the warm and the cold, the bright and the dark, etc.; and 7 (2) A multitude of innumerable Seeds in no way like each other. (P. 4 above.) I omit the ‘much Earth’ for reasons already stated. Apparently the Seeds or Parts waxed and waned on the ‘Opposites’; and we may suppose that different Seeds needed different quantities of the ‘Opposites.’ One might need more of the warm than another, and more of the dark, and so forth. But there was always some, however little, of every Opposite in it. Anaxagoras says that the ‘ Opposites’ are not separated one from the other, any more than things themselves are: the Warm is not separated from the Cold nor the Cold from the Warm (fr. 8, D.). Thus in each sensible thing there are all the Seeds and all the Opposites. And I think there is every reason to suppose (cf. De wictu) that the Opposites increased and decreased (quantitatively of course) at the will of the Seeds. But the important point to notice is that, by themselves, the Opposites are not enough. That was the point on which Empedocles needed correction. His universe of Fire, Air, Water, and Earth was a universe without a backbone. This mistake the doctrine of Anaxagoras remedies by adding to the ‘ Elements ’"—or to the ‘ Opposites,’ which correspond to them— the doctrine of the Parts. Summary.—The whole universe is ordered and controlled by Mind, which alone is unmixed (fr. 12, beginning). All other (i.e., sensible) phenomena are mixtures. These are: (1) The four ‘Elements’—haphazard vavorepyias, containing all the Seeds. They are aepa. | 3 (2) Organic substances, sensible phenomena, subject to Becoming and Change. These are ordered opotopuéperat, each of which contains a pépos of every sensible phenomenon, ready to develop in favourable conditions. The material upon which the Hepn work is the Opposites, which, by themselves, would be insignificant, because they are not sufficient to produce sensible phenomena. The Opposites are a relic of the older philosophy. A. L. PEckK. Curist’s COLLEGE, CAMBRIDGE. ARISTOTLE AND THE XOPIZMOz OF PLATO. ‘ALL the difficulties in the theory of forms arise from their separation.’ This recurrent criticism of Aristotle’s is, of course, one of the principal obstacles in the way of any reconstruction of the Platonic metaphysic. To begin with, it is flatly denied by Plato himself in the use of such words as péOecEis, mapovoia and xowwvia. It must also be rejected by the orthodox account of the Forms which takes them to be immanent, constitutive principles in the world of everyday life, like a Law of Nature or the Concrete Universal of Hegel. Finally, modern philosophy has made it clear that this view of universals is right and necessary if thought and language are to exist, and it is therefore tempting to attribute it to Plato. Relying, then, on the wé@eE:s terminology and on the necessity of imma- nence for any satisfactory logical theory, all recent Platonists of the orthodox school — Bosanquet, Stewart, Nettleship, Natorp'—feel able to disregard Aristotle’s accusation of ywpiouos. Aristotle is so often unsympathetic and superficial in his treatment of Plato that this may well be another case. We are told that his ignorance of mathematics would easily account for the mis- understanding, or that hostility to a rival school, and a pedestrian and literal mind have tempted him to exaggerate what is merely one side of the Platonic doctrine, a side which he was temperamentally unable to appreciate. It is the object of this article to insist that the ‘ separation’ cannot be thus lightly dismissed. In the end perhaps it would be justifiable to hold that it is the only tenet we can with certainty attribute to Plato. I wish also to suggest that the other available evidence for Plato’s final view confirms Aristotle and builds up a consistent doctrine resting on separation as its basis. The evidence to which I refer consists of (a) the dialogues contemporary with Aristotle’s attendance at the Academy, (0) the Letters, (c) Proclus. We must first find out what Aristotle meant by ‘ separation.’ There can be no doubt that he meant a severance which was complete and absolute. The constant contrast of xata tay modd@v with rapa Ta modAG? is meaningless on any other hypothesis. It was no mere mental distinction, or difference of aspects separable only in thought and united in reality, for this distinction Aristotle himself believed to be possible and necessary.* Again, he is careful 1 Bosanquet, Companion, p. 257; Stewart, Plato’s Doctrine of Ideas, p. 116; Nettleship, Lectures on Plato’s Republic, p. 239; Natorp, Ideenlehve, passim, especially pp. 387 sqq., 436. 1027b 30: ‘H dialpecis (sc. of Aristotle) év davoig 2 Met. A6. 987b 8 et saep.; Am, Post, Att. kal ovdx év rois mpdyywacr; cf. Met. M2. 1077b 6 772 5. with Z8. 1033b 12, 3 An. Post. loc, cit.: Elén pév ody 4 &v re rapa WONG ovK avdayKn el dwddekts Erra, elvac wévror ev kara wodNG@v adnOes elweivy dvdyxn; Met. Eg. to point : rest on ¥ (a) ° (b) * (c) 1 The poir every on form is | every obj Now matheméz argumen not sepa theory i: Aristotle only esc that ther The exist, Is | Not only and Tay! Plato is } the real | misunde: views on when it view? | superior contradi It 1 glossed ¢ more th: whole at It would It would years a master 1: Yet this totle. C 1 Met. DN cupBavdyTc 2 Gen, et 992a 9, AI 3 Ag. 9 ovclay Kal O. paration.’ principal ysic. To words as orthodox principles Universal s view of , and it is of imma- orthodox disregard hetic and ase. We ‘the mis- nd literal Platonic cannot be hold that h also to confirms its basis. rary with “here can absolute. aningless erence of istinction is careful H &v ri rapa lvar wévror év ; Met. E4. le) év diavola {2. 1077b 6 J. D. MABBOTT 73 to point out that all the detailed criticisms he has to urge against the theory rest on ywpiopos.' For instance— (a) The forms cannot cause motion in objects if separate from them.’ (b) They cannot give ovcia to objects if separate from them.* (c) They cannot make objects intelligible if separate from them.‘ The point to be stressed here is that, if there is any immanence whatever, every one of these criticisms breaks down. It is no help to suggest that the form is both immanent and transcendent, for that would blunt the edge of every objection Aristotle has to make. Now it may be abundantly clear that on occasion—and especially on mathematical occasions—Aristotle can and does do less than justice to Plato’s argument. Surely, however, such treatment is out of place here. If Plato did not separate entirely forms from things, every word Aristotle has to say on the theory is worthless. However much this would delight the detractors of Aristotle, we must dash their hopes by adding that the Platonic doctrine can only escape Aristotle’s criticism by adopting his metaphysic. It then appears that there is no difference between Plato and Aristotle on this point at all.5 Then it follows that Aristotle is exaggerating a difference which does not exist, is lacking in sympathy towards his own view, or failing to understand it. Not only this, but our only direct evidence about Plato is discarded. Burnet and Taylor have insisted with great cogency that the only safe approach to Plato is through Aristotle. If Aristotle is worthless, our voyage in search of the real Plato is checked at the outset. If after twenty years’ teaching he can misunderstand his master on a point of such central importance, what do his views on the Dyad matter? Or how could he dare to make these statements, when it must have been common knowledge at Athens what was Plato’s real view? If there were any of Plato’s pupils who did understand him—and their superior ability has not saved their names from oblivion—they would have contradicted Aristotle at once. It must also be urged that misrepresentation on this scale cannot be glossed over as lack of sympathy or exaggeration. It would show something more than ‘lack of sympathy’ if a contemporary of Kant were to base his whole attack on that philosopher on an assertion that Kant was a hedonist. It would, indeed, be a ‘ superficial’ view of Hegel that made him a materialist. It would be a curious type of exaggeration if one who had been for twenty years a pupil of William James should say: ‘ My only quarrel with my old master is due to his pathetic faith in the verbal inspiration of Mr. Bradley.’ Yet this is the position we are asked to accept with regard to Plato and Aris- totle. Can we really hold that we understand Plato well enough to be certain 1 Met. Mg, 1086b 6: Td 5é xwpifew alriov rév cupBavivtwv dvoxepav mepi Tas liéas éoriv. 2 Gen, et Corr. Bg. 335b 12; Met. Ag. ggtb 4, 992a 9, Ato. 1075a 24, M5. 1079b 12. 3 Ag. ggib 1: "Er. dd’varov elvac xwpls rh odclay xal od % ovcla; cf. Z8. 1033b 27, Mrz. 1079b 37. £ A. ogia 12: Ovdre mpds tiv émoripny odOer Bonded trav Adr\wv. . 1078b 34. 5 As Stewart admits (Doctrine, p. 111). 6 E.g. Thales to Plato, p. 312. . BH évumdpxovTa; M4. F 74 J. D. MABBOTT Aristotle could not understand him at all? If so, on what evidence can such a claim be based? Only on certain admittedly dramatic, popular and exoteric works, about whose significance there are almost as many views as inter- preters, and which may well turn out in the end to uphold the authority of Aristotle, and confirm the ‘ separation’ they are claimed to destroy. This is the point of our assertion that ywpiopos is the only doctrine we can with certainty attribute to Plato. Socrates and Pythagoras may divide the dialogues between them; the Letters may return to their old position of doubt and disgrace. If Aristotle fails us too, we cannot complain of popular or partial exposition, we cannot blame dramatic form. Aristotle is lecturing, and lectur- ing to students, and his every reference to the Platonic theory is a direct or indirect assertion of ywpicpos. If we are to use the Dialogues to check Aristotle we must, of course, concentrate on those contemporaneous with his attendance at the Academy : Theaetetus, Parmenides, Sophist, Politicus, Timaeus, Philebus. It is surely difficult to hold of these ‘ professorial’ dialogues that they do not represent Plato’s own views. As regards his earlier work such a position is most plausible, but after the Theaetetus the dramatic interest vanishes. There is no characterization and no real discussion. Each dialogue has a central figure who is not a living being but doctrine personified. Plato’s apology (Polit. 286) for the suppression of literary beauty is surely comprehensible only in view of the victory of philosophy over art—the victory so bitterly signalized in Repub. X. If the Parmenides is really Eleatic and the Timaeus Pythagorean, how is it that the dialogues grow out of each other so neatly, each using its predecessor’s results so that the Timaeus can sum up the whole, and the Philebus apply the doctrine to a special problem—that of pleasure ? Two points must first be noted. The existence of the Forms goes unquestioned throughout the whole period. Nowhere do the offending words péOckis, mapovoia and xowwvia appear, in reference to the relation of form to particular, except in the crushing condemnation of pé0e&us in the Parmenides. (xowwwvia is, of course, regularly used of the relation of forms to each other, one of the central problems of this period.) The system expounded in the Timaeus in which the group culminates rests on ‘separation,’ and would be unintelligible without it. It agrees entirely with the views attributed to Plato by Proclus and by Aristotle. It is confirmed on a most important point by Ep. VII. The Timaeus states this ywpiopos in the most uncompromising language, and two phrases seem expressly aimed against pwé0eEus and tapovcia respectively.* Its view of the relation of the form to the particular is that of original to copy.” Is this conceivable without separation? Yet it may be objected that Plato also says that the ‘ receptacle’ of determinate being is ‘ unshaped by any vontov kai dei xara rabra bv, ulunua dé wapadely- Maros OevTEpor, yéveow Exov Kai dparév. 1 SIE: ore els éaurd eicdexduevov Go AdAoGev ore avré eis AXXO mo idv Of the Form. 2 48E: év pév ws wapadeiyparos eldos irorebév, of those suggest t that the saying tl shape of 50C; the the eterr peTakv O tion? |] the ‘ goc To ayalo Plato’s ¢ him fina discover Ta pabn; minate 1 doctrine to the tr nent, as The however fair to system ¢ attribute with it. between means O and the to contr shapes ( which a: forms.’ 1 50D: | dcas wédXor 2 Ar. M jmowvcw POaprots. 3 Space palverat 5€ kai é&.6vra QuTav ... Polit. 284 Form. 4 Met. A h pirocopi the last de selves bec yap aprO poi 5 Cf, hi = can such nd exoteric 'S as inter- uthority of y- This is e can with ie dialogues doubt and r or partial and lectur- a direct or of course, Academy : ely difficult -lato’s own >, but after cterization ot a living uppression victory of XK. If the t that the r’s results e doctrine rms goes ling words of form to armentides. ich other, led in the would be d to Plato point by promising Tapovoia riginal to scted that ed by any dé mapadely- ARISTOTLE AND THE X0OPIZMOS OF PLATO 75 of those forms which it is about to receive. This appears once more to suggest that creation makes the forms immanent. It is quite clear, however, that the déas of this passage are not Forms at all. Plato is not content with saying that the particular resembles the Form. Itcan only do so by having a shape of its own.” It is these shapes which are the eio.dvta xal éfiovta of 50cC; they are wrunuata Tovtwr (i.e. of the true Forms) which mediate between the eternal, unchanging, separate eddos and the formless ywpa.* They are ra petaév of Aristotle’s account. How can ra petaév exist if there is no separa- tion? In the view of the Politicus the indeterminate dyad of human action, the ‘ good-and-evil,’ is determined by receiving a copy, always imperfect, of To aya0ov; the dyad of aesthetic form is determined in the likeness of to xadov. Plato’s continuously increasing interest in mathematics and Pythagoreanism led him finally to hope that all determination of objects might in the last resort be discovered to rest on mathematical relations.‘ In that case ra petaéi become ra pabnparixd ; all dyads are reducible to one, To wéya Kai prxpdrv, the indeter- minate in size and shape, ywpa, space. It is true that the Timaeus puts this doctrine in the figurative form of a creation myth; but even a myth is a guide to the truth. The Timaeus has just as little meaning if the Forms are imma- nent, as the Republic myth would have if the soul were mortal. The evidence of Proclus, who inherited the esoteric Academy tradition at however distant a date, confirms this account throughout. It would be hardly fair to quote his commentary on the Tzmaeus, which naturally expounds the system as we have given it. On the Republic and the Parmenides, however, he attributes to Plato just the same system, and tries to make the Republic square with it. The central point, as we have insisted above, is a clear-cut distinction between the immanent ‘ shape’ and the transcendent or true form. It is by means of this distinction that Proclus reconciles the separation he and Aristotle and the Timaeus assert with such a passage as Rep. 476, which seems flatly to contradict him.® He says that Socrates confined himself to the immanent shapes (ra eiotovta Kai é€vovtra of the Timaeus) but later passed on from these, which are changing, transitory and imperfect, to the true and separate eternal forms.? The Parmenides Commentary insists on just the same distinction.® 1 sop: duopgov dv éxeivwy admacay trav idedy, daas wédroe Séxer Oat. 2 Ar. Met. Z14. 1040b 30: [The Platonists] 6 xowwvia mpdtewv xal cwudtwv Kai ddAjAwv. 7 In Remp. 422: pnréov airov . . . dpxerOar per amd Tav perexouévww elddv . . . TedevTav Je els Ta mowlcw «SC. Tas tdéasm Tas abTas Tw elder Tois Pbaprots. 3 Space is dacynuarifomevov iro Tay eiordvTwr. palveras dé 5c’ éxeiva Gore GANoiov. 7a 5’ eiordvTa kai éiévra TOv SyTwv del piujpara, TUTwWOévTa am’ airav . .., etc. [50c]. Cf. with the pérpov of Polit. 284 sqq., which of course cannot be the Form. 4 Met, Ag. 9934 32: yéyove Ta wabjuara Tois viv h girocogia. This of course ied straight on to the last development by which the Forms them- selves became numbers. De An. 404b 26: of wév yap dpOuol ra el5n atta . . . EXEyorTo. 5 Cf, his claim in Theol, Plat. I. i. auebexra tav elddv & 5 éort mavTedds SvTa xal woavTws éxovTa. 8 In Parm, II. 167, Stallbaum, p. 573: E.g. mpd0ecis pev Hv TO XwKpdry perayayelv dmwd Tov kaTareraypuévwv povddwy thy Snrnow én’ abtrds ras cEnpnudvas xai duepets airias; In Parm, II. 118, 119, St., Pp. 543, 544: wav dé mwd7HOos Exew évdda Sirrhv, Thy wéev cuvreraypevny, Thy Se e&npnuéevyy. idé yap aird mpGrov émi trav etddv, olov Srws 6 &vOpwros dirrés, 6 wev eEnpnudévos, o dé werexdpuevos. kal $rws 7d kadov dirrév, Td wev Wpd TSv WorAAGr, TO dé é€v Tots wodKoits. . .. €ore d€ ad’ Trav wordy decuds 7d €v adrots kowdv, kal dia TolTo Gos pev O ard dv@pwmos, &dAos dé 6 €v Tots KaGéxaora. 76 J. D. MABBOTT The same point is also made clear in Ep. VII. There the form of circle is sharply distinguished from the ‘ shape’ of circular objects. The immanent shape is (a) imperfect, (6) transitory ; the Form is perfect and changeless. It is the ultimate goal of knowledge, but we cannot come to know it from study of the sensible particulars.‘ Philosophers who find this passage in disagree- ment with their views on the immanence of the form have cast doubt on the canonicity of the Letters. The close parallel between Ef. VII. and Proclus might well suggest that they are neo-Platonic forgeries, foisting on Plato neo- Platonic doctrine. Forgeries of course they must be, since it is possible that Aristotle and Proclus should misunderstand Plato, but hardly that Plato should misunderstand himself. Any such theory, however, is shattered by the fact that the Letters were known to Cicero. We find ourselves therefore forced to push back neo-Platonic doctrine to a time before neo-Platonists existed, and it is surely possible to suggest that after all the Letters may be Plato’s writings, and if they inculcate neo-Platonism, then Plato held that doctrine, and Plotinus and Proclus, sole trustees of the Academy tradition and the aypada Soypata, are to be given more credence than in the past. In any case it is only the philosopher, with an a priori prejudice as to what Plato ought to say, who condemns the Letters. All the great scholars, free from philosophical bias— Bentley, Meyer, Burnet—have held them to be genuine. If we reach the conclusion that every scrap of evidence for Plato’s final theory confirms the unassailable assertion of Aristotle, we are still left with the original difficulty. What are we to say of pé0eks, rapovoia, and Kxowwwvia ? The explanation of Aristotle is that this terminology, taken in conjunction with the vrapddevypa view of the later theory, is nonsense ;? and so it would be if Plato ever urged it of his final doctrine. Mr. Ross* holds that, since Aristotle recognizes péGcEis as a Platonic term, he must have taken the Platonic forms to be both immanent and transcendent. We have insisted sufficiently above on the difficulty of such a view. Such forms could not be called ywpiord, and even if they could all the arguments which rest on ywpiouos would break down. If the form is immanent and transcendent, it can give motion, being, and intelligibility to its objects. Aristotle, holding that the form is entirely transcendent, is compelled to call wé#e~s a poetic metaphor and an empty word. Is it not possible, however, that wé@eEvs and mapovcia are survivals of the real Socrates? Nowhere is Burnet’s theory more convincing than in dealing with the Theory of Forms, which Socrates always introduces as well known to his disciples, and which they receive without further question. Again, as 1 342c 1: Tplrov dé rd Swypadovuevow re Kal Tis évavrias tye év airy dicews. Cf. with Phaedo éEarer@duevoy kai Topvevduevov xal dro\\iuevov Gv 74b sqq., where it is shown that knowledge of avroés 6 KUKNos . . . obdév mdcxe ToUTwWY ws ETepov avrd Td tcov can never come from study of 7a ica bv; cf. 343a 5: KUxdos Exacros trav év tais mpdtect since it is other than they. ypapoudvwy xal ropvevOdvrwy peords Tod évavTiou 2 Met. M5. 1079b 25: 7d Aéyeww wapadelyuara éorlvy Tw wéumrw (the separate form)... adrds adra elvar cai ueréxew airav raidda Kevodoyelv éore. d€, payev, 6 KUKXos obre TL oOuixpdrepov obre perfor 3 Aristotle, Metaphysics I. xlii. Burnet | History Plato, c everythit If Socra immaner must be irrelevan If t reasons ° made th as a pu] its most of the v must hav him to tl the a pri a perish element the univ mysticis! an escap reality.* experien one dial of knowl gives it 1 Ins it solves The dial qualities expound way In \ understz (Ideal) * yw pla LO only the that it counter] ‘Burne 2 Cf. Xe tls eln arya karéNutre 6¢ 3 Ar. M 4 E.g. a of circle immanent reless. It rom study disagree- bt on the d Proclus Plato neo- ssible that ato should y the fact » forced to ted, and it 3 writings, d Plotinus , OOYmaTa, s only the say, who -al bias— ito’s final t with the colvwvia ? ‘tion with wuld be if Aristotle nic forms tly above oTd, and ld break n, being, entirely n empty Is of the 1 dealing nown to vain, as ith Phaedo wwledge of y of ra ica padelypara orvyety ore. ARISTOTLE AND THE XOPIZMO® OF PLATO 77 Burnet points out, the theory comes as the culmination of the ‘ Intellectual History’ in Phaedo 96 sqq. to solve difficulties which only Socrates, and not Plato, could have felt, as the goal of a philosophical pilgrimage which fits everything we know of Socrates and contradicts everything we know of Plato. If Socrates did originate the Theory of Forms, his forms must have been immanent, for immanence is essential to the Socratic methods. The form must be discoverable by comparison of examples and careful elimination of the irrelevant.? If there is anything in this suggestion, it is not difficult to see several reasons why Plato should have taken these immanent forms of Socrates and made them transcendent. In the first place Plato began to study philosophy as a pupil of Cratylus,? and this early acquaintance with the ‘ flux-theory ’ in its most destructive form might well account for a wish to keep the forms out of the world of sense and change. Secondly, Pythagoreanism, which Plato must have studied with deep interest on his visits to Sicily, would introduce him to the mathematical view, with its tendency to set thought against sense, the a priori against the empirical, and to regard the visible as never more than a perishing and imperfect copy of the intelligible. Thirdly, there is the element of mysticism—not the nature mysticism of Wordsworth, which sees the universal shining through the visible (as Stewart holds), but the abstraction mysticism of the East, for which the propaedeutic is mathematics, and the aim an escape from becoming into the other world, the world of pure and timeless reality. Against the view that Socrates, and not Plato, was the mystic whose experiences are referred to in the dialogues, we may set the fact that, in the one dialogue where the mystical vision seems expressly indicated as the goal of knowledge, Plato carefully refuses to allow Socrates to claim the vision, and gives it instead to Diotima of Mantinea. In support of the development view suggested above, we may examine how it solves another contradiction besides that between ywpiopos and pébekcs. The dialogues tell us both that there are and that there are not Forms of evil qualities. The Euthyphro, a dialogue usually admitted to be ‘ Socratic,’ expounds the Theory of Forms in 5D, 6p. In view of the perfectly explicit way in which all the usual points about the theory are stated, I am unable to understand Adam’s note (his edition, p. 65): ‘ There is no reference here to the (Ideal) Theory as yet unborn.’ It seems to me that all that is missing is the ywpicuos, which we take to characterize the later theory. If, as Adam holds, only the Socratic Aoyor are in question here, such a view is possible, provided that it is made clear that the Aoyor are not definitions, but their real counterpart, not thoughts or class-concepts having only a subjective or mind- ‘Burnet, Phaedo, pp. 95, 96. of them, it appears, use mathematics to train 2 Cf. Xen. Mem, III. ii. 4: xal otrws émicxomGv them to rise to the contemplation of pure universals. Cf. The Travel Diary of a Philosopher I. 267, where Keyserling explains how he himself achieved this result. ris etn dryadod tryeubvos dperh, Ta ev Ga wepippec KkaréXurre 5¢ 7d eddaluovas moveiy Gv aviynra. 3 Ar. Met. A6. 9874. 4 E.g. among certain sects of the Yoga. Some 78 J. D. MABBOTT dependent reality, but the real constitutive characters of real things. ‘viding, therefore, that the Theory of Forms is expounded here, I wish to insist that Plato attributes aform to To dvootov. Here we have forms of evils associated with the immanence view (€v tdaon mpdéer). The same coincidence occurs in Republic 475 sqq.2. The unity of the evil quality is just as strongly asserted as that of the good. Now to one who holds an immanence theory, and who is also, like Socrates, interested primarily in Ethics,® the existence of forms of evil qualities is quite a natural assumption. The moralist always prefers to believe in the independent reality of evil, and a method limited to the analysis of particular actions would naturally conclude that cowardice has just as positive and definable a nature as courage. If, however, the forms are separated from things, and abide in a perfect and timeless world of their own, there is surely no place in that world for the forms of all the evils, for perfect ugliness and absolute inequality and eternal injustice. In the final position, represented by the Professorial dialogues, Plato realizes this quite clearly. An evil action or ugly object is not an imperfect copy, in which the Demiurge has tried to reproduce perfect evil or ugliness. It is an action which falls short of the perfect beauty or goodness. Evil is found only in the world below, the world of time and change. This point is reached by the Theaetetus.° There is a further question: Why did the Demiurge fail to make his copies of the form adequate? The answer to that is the final advance achieved by the Polzticus. Since we cannot attribute to the Demiurge a wish to produce the imperfect, we must hold that his failures are due to refractory material.© This may not be an altogether satisfying account of the status and origin of evil, but it is a logical and consistent view, and one which has won supporters throughout the whole history of theology. Proclus deals with the passages in the Republic by the same method as he used to account for xowwwvia, by saying that here Socrates has not advanced to the distinction between the immanent ‘shape’ and the transcendent form. In the sensible world there is indeed good and evil, and either may be mentally abstracted from its particular setting. In the world of true and separate forms there is good only, and in a final analysis evil must be treated as 1 5d: wotév ti 7d edoeBes dys elvar kal 7d doeBes Kai wept pévou kai wepi Trav GN\Awv ; 4H ob Tavrév éortiv év wadon mpdte 7d So.ov aire aire, Kal 7d dvdcrov ad rot per dciov mavTos évavriov, abrd 5é aire dpuotov kal €xov play twa ldéay xara rHv dovérynra ray, 6 ri wep av péd\Ay avdcrov elva (do1dryra Of MS. B is preferred by Adam to dvooisrnra of the other MSS., which would make xard easier to trans- late and also strengthen my point). Cf. 6p: ... €xeivo abro 76 eldos, @ wavra Td down bord EoTw. 2 476A: xadov kal aicxpév . . . év éxdrepov.... kal mwepi 5% dixaiov xai adixov Kal dya0od Kal Kaxod kal wadvrwy Tay elddv mépt 6 abros Nbyos, a’ré wev ev éxacrov elvar, Tn 5¢ Tov mpodtewy cal cwudrwv Kal G\AjrAwy Kowwvia wavraxod davraféueva moda palvecbar Exacrov. 3 Ar. Met. A6. 987b 2; Xen. Mem. I. i. 10, 16. 4 Cf. Phaedrus 2508: Suatoctyng piv ody cai owhpoctync Kai boa dda ripua Wvyaic. 5 176A: ovr’ amodicOa ra Kaxa dvvariy, & Oeddwpt,—vmevayriov yap te ry ayabye iva avaykn—ovr’ ty Oeoicg adra wWpiacba, rny Ct Ovnriy gio cai Tove roy romoy mepiTroXEl E avayene. 6 273B: mapa piv yap rov ovyrévrog (i.e. dnp- ovpyou—Tim. ; airiag rij¢ pitewc—Phil.) ravra Kad Kéxtnrat* wapa ¢é Trig EumpooOev eEEwe doa xarera wai ddua ty oipav@m yiyvera, ravra é ixeivng abrog re Exe Kai Toic Zyoig amwepyazerat. deficiency. distinctior ‘separate as a chal (vaTEporyer It is dogmatic. dogmatic that the « and Proc: Nettleshi 1 In Ren dpxyecOar pi redevrav OF ovKETL ELV cewy* ov ¥ orephoec T TACEC. » - perexopevn 2 In Rem elvat, TO pe aisxpov tv héyery * et « oiding, insist that associated - occurs in ly asserted ind who is f forms of ys prefers ed to the e has just a perfect ‘Id for the 1d eternal es, Plato imperfect ugliness. ss. Evil S point is did the r to that ribute to s failures atisfying ent view, heology. he used -d to the rm. In nentally separate sated as I, i. Io, , ed ‘ ‘VY OVY Kat UVATOY, w ab sivar 6& Ovnriy vayKNC. (i.e, Onp- l.) wavra éEewe boa ravra é& wera. ARISTOTLE AND THE XOPISMOS OF PLATO 79 deficiency.! This view, Proclus tells us, is held by Plato only in virtue of the distinction he later makes. ‘separate’ form, he has to admit that there are forms of evil. While he does not distinguish ‘shared’ from For evil exists as a character of things and also as a possible abstraction from things (Sarepoyeves vonua—emi Tois mroAXois), but it cannot exist as a separate form. It is inevitable that such an account as the above should be cursory anc dogmatic, but any account of Plato must always be more or less unwarrantably dogmatic; and this paper is written in the belief that it is better on the whole that the dogmas used to supplement the dialogues should be those of Aristotle and Proclus than that they should be borrowed—as in the work of Bosanquet, Nettleship, and the Marburg school—from Kant and Hegel. 1 In Rempub. 422, 423: pnréov avrov... dpxyecOar piv ard rev perexopivwy eiddv... redevray ob sic ra duéOexra ry cidGv. . .. 0 Kai obkére peuvijcOae Tov aicxpod cai bdwe THY oTEPN- cewy* ov yap eiow ty roicg apeBexrog eideow at orephoac ray eidov. ob: at Tey KaKiy Urod- racec.... Kowd ra Eeyopeva arayTwy, Kai Toy perexopévwy kai Tov aueGixrwr. 2 In Rempub. 423: ei piv ody dvo ravra édeyev elvat, TO peptkdy Kai rH idtav, nvaykalero Kai TO aicxpov ty Te Tapa Ta TOA TiOEig Wéeay aicxpod héyery * ei dé way eldog avridiaipei Tpdg TO peEpiKoY, J. D. MAssBotrt. ovdév av éyou GAO 7 Ort éoriv Kai aisxypdy eiddg rt tv * rovro dé ov wavrwe idéay selva dvayedaZea, ad’ ty roic xaQ’ teacra Kara arérrwow Vrocray Tij¢ idéac, cai eldog AEydpevoy we ToAROIC KOWdy, H év raic Wyaig we vonua vorepoyevic, obk ayvwiev airaig éyyevdpevoy, GAdd carwOev. 2& wy dijdov Ort Tad piv Kara pvow rpiywe tori, TPO TwY TOG, ty roic mrodXoig, tri roig modXoig, Ta O& Tapa giow dtyace tv roig modXoig wai éxi roig woddoic. The added distinction of the abstracted concept from the immanent ‘shape’ also occurs in £#, VII. 342e sqq. THE RAESVS. As my suggestions have been unfavourably criticized in Professor Ridge- way’s article on ‘ Euripides in Macedon’ which appears in the current number of the Classical Quarterly, perhaps I may be permitted a few words of explana- tion and reply. Professor Ridgeway’s article in its latter part is chiefly directed against the thesis I endeavoured to support in the Classical Review of 1921, pp. 52 sqq., to the effect that our Rhesus is probably not the genuine work of Euripides. Space does not permit me to do more than refer to a few of Professor Ridgeway’s strictures; but I would ask readers not to accept his statements of my views without reference to my original article. The context often makes all the difference. Thus, for example, when I am arguing that a strong case can be made against Euripides on the score of language, I am made to be dealing with the whole issue of authenticity. Indeed, if Professor Ridge- way’s judges were to sum up at nist prius on the lines suggested for them, they would soon become the by-words of their profession. (1) The object of my calculations, of which Sir William thinks but little, was to contrast the style of the Rhesus with that of a genuine play, and the A lcestis was chosen partly to meet Professor Murray’s view and partly because of its date. The choice was made against my own interest ; and if I had taken, say, the Orestes or the Bacchae, 1am confident that the results would have been even more striking. But this doesn’t matter, since Professor Ridgeway attaches no importance to differences of style, which he brushes aside by averring that Euripides changed his style (even in the use of 6s and éyw) in accordance with his subject. Even if that were so, I do not think that I should myself describe the Alcestis as a semi-comic play. (2) In the matter of verbal repetitions, I expressly guarded myself against being supposed to refer to repetitions merely careless, like that of Jon 2. If the list of repetitions that I gave does not appeal to Sir William as remarkable and indicative of a poverty of talent, there is no more to be said. (3) I am charged with giving away my own case because I allow some merit to the Rhesus. But my case is that, good or bad, our Rhesus was not written by Euripides, and I maintain that neither its merits nor its faults were characteristic of him. (4) Sir William Ridgeway blames me for making no attempt to show how a spurious Rhesus replaced the genuine one in the MSS. The allusion to the mediaeval MSS. is nothing to the point: the play was there because they are descended from the Alexandrian édad¢y. I hoped that I had made my position absolutely clear in this regard. As I said, it is impossible to give any answer to the c But it i current, Alexand refers tc p. 148 h in the 1 suggeste Neverth actually Alexand and the generall product: 1 There reference (C.R. XX Dr. Alcestis Judge \ another ous for been su adoptin of mine when t ‘semi-c terms ‘ of ‘ care revised Euripic was jus by argu not dec play, h not nez the lite fine, an charges ssor Ridge- 2=nt number of explana- fly directed ‘w of Ig2I, luine work o a few of accept his he context ling that a Iam made sor Ridge- hem, they but little, y, and the y because iad taken, have been y attaches tring that ince with ‘ describe f against 2. Ifthe narkable Ow some was not ts faults iow how 1 to the they are position answer THE RHESVS 81 to the question how the spurious Rhesus came to be included in the canon. But it is certain that it was much suspected, that two extra prologues were current, and yet that, notwithstanding these peculiarities, it was included in the Alexandrian edition—presumably that of Aristophanes. Professor Ridgeway refers to his defence of the play in his Origin of Tragedy (pp. 147-150). On p. 148 he says it is difficult to see how the real Rhesus could have been replaced in the many MSS. of his works—or, rather, in the Alexandrian canon, as suggested above—by an inferior and spurious play on the same subject. Nevertheless, something like what Sir William regards as incredible has actually occurred elsewhere. Euripides wrote a Sisyphus,’ produced with the Alexander, Palamedes, and Troades in 415. Nearly every trace of it was lost, and the fragments quoted from Euripides by Galen and some others are now generally ascribed, in agreement with Sextus Empiricus, to the homonymous production of Critias. A. C. PEARSON. omission of the words ‘the tetralogy,’ and in any case matters little for the present purpose. See also Wilamowitz, Anal. Eur., p. 166. 1 There is an unfortunate blunder in the reference to this matter in my previous article (C.R. XXXV. 61). It can be cured by the REJOINDER. Dr. Pearson from the relative frequency of thirty-three small words in Alcestis and Rhesus said that he had a case to go to the jury. I replied that the Judge would point out (i.) that there was no evidence for the existence of another Rhesus, {ii.) no motive had been shown for the substitution of a spuri- ous for a genuine work, (iii.) nor had any machinery for effecting such a fraud been suggested. Dr. Pearson still does not meet any of my objections. His adopting the Alcestis as a test to meet Professor Murray’s theory is no concern of mine. The difference in the relative occurrence of ds, éy# is of little value when the subject is so different. He would not describe the Alcestis as ‘semi-comic,’ but the Hypothesis says: copscwrépay exer THY KatacKeuvnv. He terms ‘careless’ Qeav twice in Ion 2, but why should there not be instances of ‘carelessness’ in Rhesus, written at the end of the poet’s life, before he had revised it? Dr. Pearson does give his case away, for he holds that either Euripides wrote the Rhesus or its author copied him. He admits that Eysert was justified in saying that the genuineness of the play could not be disproved by arguments based on language; he admits that ‘ dissection of phraseology is not decisive,’ and finally after trying to show weakness in the structure of the play, he says that ‘the curious thing is that when all has been said the play is not nearly so bad as it ought to be’; he admits the ‘skill’ of the plot, that the literary workmanship is by no means ‘contemptible,’ that the lyrics are fine, and some speeches very effective; and finally he admits that some of his charges might be made against genuine plays of the poet. WILLIAM RIDGEWAY. THE (HOMERIC) HYMN TO HERMES. 274 « 8 eOéXdes, rarpds Kepadny péyav OpKov dpovpas * py pev eyo pyr’ adrds trioyopas aitios elvas pyre tiv’ GAAov Orwra Boov KXoTbv vperepdwv, aitives ai Bdes etoi* Td 5€ KA€Os Olov dxovw. ws ap’ pn kal muxvov amd BAehdpwv apapioowv oppiar pirrdferkov dpwpevos €vOa Kai évOa, pap’ arorvpi(wy adcov Tov pvOov aKovwr. IF all or any of our MSS. dated from 800 or goo B.c., it might be of importance to note GéAes for the regular epic form é0éAes, and even to print it so in the text, otherwise it is negligible. More worthy of attention is the punctuation after dpotpac. The presence of imicyowa: in the next line is held to justify the stop given above, otherwise the comma, as in Gemoll’s edition, would be sufficient or more than sufficient. For in accordance with Homeric usage (cf. A 76, Y 313-5, K 321, etc.) we might read : TaTpds Kepadny péyav Gpkov opovpat py pev eyo pnt’ adrds—irioyopa:—aitcos eivar. Hermes says, ‘ I will swear—I promise you ’—instead of the more direct ‘I undertake to make oath.’ This I submit is the plain meaning of the passage, obscured a little, but not seriously, by the position of iaicxouac necessitated by the metre. In 277 Mr. Allen is undoubtedly right in reading ai tives for the usual aitives (Gemoll, Abel), for it is in the old epic rather than in any later writing that the separate writing of the component parts of édor1s should appear. But what of ai Pdes? Is it ‘those oxen,’ the particular ones you have been speaking of? Emphatically no. The rest of the line makes it absolutely certain that Hermes was disavowing any knowledge of oxen except from hearsay. The writer of ai oes meant the article to be generic as in oi dv@pwro, used to express ‘ mankind’ in general, one of the latest of later usages, and one that cannot possibly be attributed to the writer of this Hymn. To him oxen in general were Boes and ‘ mankind’ av@pwro without further addition. The meddler thought he knew better, and brought in ai to prove his superiority. We cannot blame him, but we can remove his damaging handiwork and restore in accordance with 311: ai Tivés eior Boes* 7d Se Sx) KAEOS olov aKxovtw, It is not surprising that many editors have doubted the correctness of oppvc+ pirtafecxev from Ruhnken downwards. It is futile to imagine that the expression is in any degree elucidated or defended, as S. and A, seem to think, by the fact that piztw is found in the sense of jirrw éuavrdv. The baby Hermes has many apelike tricks, but he could not, like the monkey celebrated by C. S. Calverley, ‘ hang By his teeth or tail or eyelid To the slippery mango-tree.’ opps purrdferxev (Hermann) is usually read, and some such change is certainly needed. I cannot but suspect the double-shotted iterative, and would suggest # 9 > , oppy avappirracKey, ‘he kept throwing up his (two) eyebrows.’ The form furracxe occurs at 9 374, d 592, al from the The at emen« GAvov ws. best MS I have ir should fi ‘and eve (i.e. false aKOV recogniz word,’ al No account plagiaris eyes twi beads, m giving a stood wi solemnit from the The the pre: expressil believe . It 1 Oxontens but in 7 dvTeropet The because necessar earlier, of expre is, the c even mc S.¢ stripped Apollo r this rem The we find ava oko it is aft sense W importance n the text, ter opovpac, iven above, more than < 321, etc.) | undertake red a little, sual aitives g that the it what of aking of ? ermes was Boes meant eneral, one the writer -o. without i to prove handiwork of ddpior pression is » fact that ny apelike ang certainly rest at 9 374, THE (HOMERIC) HYMN TO HERMES 83 A 592, also ¥ 827. The spelling should be jizrecxe, but the tradition is for pirracxe, from the influence probably of furrdfw. The concluding line is by no means free from difficulty, as the various attempts at emendation show: trocywv, iroords, aviccwv, tAaxtav for dxovwv ; avidv@ ws for dAvov ws. In regard to this last it is noteworthy that «ds is the reading, not rdv, of the best MS. M, and yet most of the later editors exclude it from the text, and, as I have indicated, show dissatisfaction with dxotwv. I would rather suggest that we should find accommodation for ws and acquiesce in dxovwy, e.g. paxpa 5’ exvpil’, as aAidv Tuva pvOov axovwv, ‘and ever and anon he gave a long whistle, as does the man who hears an idle (i.e. false) tale.’ axovwv is the epic form of the later 6 dxovwv, though many are chary of recognizing so elementary a fact. It follows that rdv piGov could not mean ‘ that word,’ and accordingly rva is a necessary correction. No better illustration of this passage could be found than Mrs, H. B. Stowe’s account of Topsy in Uncle Tom’s Cabin. The verbal resemblances, where no plagiarism can be suspected, are striking. ‘‘* Yes, mas’r,” said Topsy, her wicked eyes twinkling as she spoke.’ Again, ‘ her round, shining eyes, glittering as glass beads, moved with quick and restless glances over everything in the room.’ ‘ Finally, giving a prolonged closing note, as odd and unearthly as that of a steam-whistle, she stood with her hands folded and a most sanctimonious expression of meekness and solemnity over her face, only broken by the cunning glances which she shot askance from the corners of her eyes.’ The long whistle of Hermes is, of course, not the same as Topsy’s. It is still at the present day the recognized method, except among the cultured classes, of expressing astonished incredulity. Hermes is pretending to be quite unable to believe Apollo’s tale about the oxen. 282 @& rerov, iTeporevta, SoAodppadés, 7} we y’ Oiw ToAXakis avTiropevvta Sdpous €d vareTaovTas EvvUXoV OvyX Eva pouvoy em’ ovde Hota Kabiooat oKevafovta Kat’ oikov arep Wodov, of dyopeves, It must be a misprint that gives us SoAoppadée in the Scrip. Class. Bibliotheca Oxoniensis. For 4 oe pad’ olw either 7 cé y’ dé, as in 156, or 9 Kéo’ diw is required ; but in this form of expression the aor. xa@icoo. without xe is quite legitimate. avriropevvta is Hermann’s correction of avriropovvra,. The explanations of the origin of éx’ ovdei dora xabiooat seem unsatisfactory, because it is generally, I might say always, assumed that the attitude described is necessarily associated with robbery. From the time of Job, and probably much earlier, ‘this sitting on the floor’ has been in the Orient the recognized mode of expressing deep affliction and mourning. Robbery may be, as in this instance it is, the cause of the trouble; but, as in the case of Job and his friends, many others, even more serious, may be named. S. and A. say that ‘ Apollo regards Hermes as a practised thief, who has‘already stripped more than one house and has a wider career before him (cf. 159). Neither Apollo nor Maia (159) have any such idea; they are speaking solely of the future of this remarkable infant. The one serious defect in our passage is oxevafovta kat’ oixov. In the early epic we find only xara ofxov (s u—v), never, save here, xar’ ofkov. The obvious remedy is ava otxov, which in later Greek would be unmetrical. As for oxevafeev, common as it is afterwards, there is no trace of it or its cognates in Homer, nor can it bear the sense which has to be forced upon it here, that of ‘ plundering.’ 84 T. L. AGAR In view of Hesiod, Sc. Herc. 468, Kixvov oxevAcioavtes ar’ apwv Tedxen Kadd, it seems extremely probable that cxvAcvovr’ should be read here, either in the general sense of thieving or, even more precisely, of removing armour and weapons stored as valuable and serviceable property. oxvAdw is found Anthol. Pal. 3, 6; hence we might read oxvAd(ov7’ here as closer to the tradition (cf. dryuaw and atiudfw in Homer). Only slightly less likely would be cvActovr’ (Il. V. 48, XXIV. 436) or cvAdgovr’ (Hes. Sc. H. 480) Bin ctrAacxe Soxedwv, and Il. 1V. 116 aizap 6 cia ropa papéerpys. € / “A OTOTaY KpEeLov Epat iCwv 287 > ~ , 4 > , Dt avVT)S BovkoAiouw K@l ELPOTOKOLS OLETCO LV, Read ovére xpeov or xpedwv. In 288 all the MSS. read avrijoces (avtioygs L) ayéAyot Boov kat moeor pijAwv, but on the margins of ELPT ayrnv BovxodAdouws Kat eipoToKats diecowv is found. From dvrnv Baumeister has evolved davrds and Gemoll It has been said rightly enough that both variants give exactly the same meaning, and quite wrongly that it is hard to see how one is preferable to the other. It is impossible to find any reasonable objection to any part of the vulgate reading linguistically or otherwise: érmdére—dvrijoes, where we have the true epic non- thematic form of the aor. subjunctive, is exactly parallel to A 351 > ~ QVT})S. ommor “Ayxasot Tpwoiv ef’ trroddpowww eyeipopev ofdv "Apna, I. 646, and elsewhere. The appearance of avrnv is supposed to negative avrijcess along with its later form dvrijoys. That is a very needless conclusion. If someone desired to explain the nature of the meeting, he could not do so more clearly than by adding the gloss avrnv ‘ face to face.’ On the other hand the forms dvrgs and avrys have little or no warrant in early epic, and Bovxodiow: has none at all. In 498 of this Hymn PovxoAjas is different in form and more so in meaning. Attempts to make glosses scan by hook or by crook are by no means rare. ‘a ‘ on \ , 2 , , sd 29I TOUTO Y2p OUV KQ@t €TELTA fer aGavaros YEepas EELS apxos pnAnréwv KexAnoeat Huata TaVTE PX hat oe , ° Clearly not per’ d@avarors but per’ dvOpwros. The absurdity of the tradition is undeniable. Unfortunately both readings have the same metrical defect. A ready correction would be per’ dvipecoww, but the redundancy of yap otv «ai rather favours a more extensive change, such as TOUTO TEedv yépas Eocer’ Ev avOpwroow erecta (cf. 458) apxos pnAntyot KekAHoeat Huata TavTa. ‘Thou shalt be hailed by thieves for ever as their captain.’ Of these verses the former is merely tentative. Other suggestions might easily be offered. Confirmation must in every case be lacking. So, too, in the latter verse many may prefer to transpose thus ¢7Anryo’ apxds, needlessly I think, for the shortening of the first syllable of xexAsjoea is legitimate enough, cf. IlérpoxXe (— uv) T 287. The Moscow MS. reads atyos, cf. edx7 and etxos, and this last is favoured by Ilgen; but if dpydés is to be interfered with, the alternative must be adyos. 294 ovv 5 apa ppacodpevos tore 57 Kpards ’Apyecpdvrns. There is little doubt that the pronominal article has here been ousted by the never-failing dpa. Read civ & 6 ye, and compare 365 % Tou ap’ ds eirwv with the same defect, revealed by A 68 7 ro: 6 y’ ds etrwv, A 101, and other instances. So also in 370 HAO. 6 y’ és nérepov should most certainly be read, and perhaps 94 récaov 6 das, or even T¢ For rote é This its impro] by the av end. No in the sar and ayye. the early as is quit has not | god of tl loudly on The trad support t The certain a imp’s ea Naturall: He recog should b Odporec in Tho more cle OTEVOE Tt of the re so that | ‘ pushed ears are wrapped himself t In the n The Compar MAn' the general ns stored as ; hence we ‘in Homer). or avAd(ovr’ paper pns. ‘avrnons L) KoAVowwt Kat ind Gemoll y the same 9 the other. ate reading } epic non- ve avrnoess [ff someone ‘ly than by nt in early different in rare, radition is A ready - favours a ight easily itter verse <, for the Ke (- vu uv) favoured 1S ed by the the same o also in cov © das, THE (HOMERIC) HYMN TO HERMES 85 or even téc0’ 5 ye das (Aristarchus is said to have read régo’ dydpeve B 230, o 271). For tore 53) kparis perhaps 767’ evoKoros as 73. 296 rAnpova yaortpds épiOov ardabadov dyyeAcwrny. This line should be expelled from the text by every future editor, not because of its impropriety, but because it is a manifest fabrication, never written or composed by the author of the hymn. It is a tissue of senseless pomposity from beginning to end. No possible meaning can be given to rAxpova, the first word; yaorpds épiOov is in the same predicament. Then drdac@adov is the merest nonsense in this connexion; and dayyeAwrnv, however imposing in appearance, is a form absolutely unknown to the early epic. The line is certainly spurious, even if Hermes did what is supposed, as is quite likely, and should be removed, full-stop and all. The wonder is that it has not been denounced before. Hermes, on hearing that he is to be hailed as the god of thieves, makes up his mind at once, and with all speed (€covpévws) sneezes loudly on purpose (airés), as Catullus says, ‘ sternuit approbationem’ : éeooupevus Oe pey’ avrds ererrape. The traditional per’ airév is an attempt, and a poor one, to give countenance and support to 296. 303 Tovrots oiwvoiors ot 8 ab’ dddv iyyepovetoes. The two first words refer, I should say, not to anything Hermes has done, but to certain actions of Apollo himself, who, as it appears from 305, has boxed the little imp’s ears. The buffets administered are the omens he mockingly interprets. Naturally Hermes jumps up at once, and, in common parlance, ‘ gets a move on.’ He recognized that Apollo was in a hurry éoovpevds rep dd0t0. Probably jyeudvevorov should be read for syemovedoes, and even more probably Oapcecis, orapyaviara, for Odépoe: in 301. Apollo does not say ‘ Cheer up,’ but ‘ You are impudent.’ 305 omovd; idv: apdw dé rap’ ovata xepoiv ewer, omdpyavov aud’ apouriv eeApévos, etme 5é pvOor, Though orovdy idv does not falsify the sense here, the meaning would be far more clearly conveyed by omevde 5 iwy ‘he went on with speed,’ as in the idiomatic orevoe Tovncdpevos (1 250) ‘ hedid his work with speed.’ The reading and punctuation of the rest combine to spoil the picture. Thecomma separates orapyavov from éewfet, so that Mr. Allen must finally have adopted his curious suggestion that Hermes ‘ pushed back both his ears.’ I say emphatically no to this. The wise infant, as his ears are being boxed, protects them by covering them with the orapyavov which was wrapped round his shoulders, pulling it up with both hands. Then, having protected himself to this extent, he resumes his wrangle with Apollo. Accordingly I would read: orevoe 8 wv: dudw dé rap’ ovata xepoiv ewber omrdpyavov apd’ aout éeApevov, pxe Se pvOov * In the main this is Baumeister’s reading. 310 tperépas exrAeYa Boas, ovd’ adAov drwra, The iambus in fourth place is due to an omission. Read for the metre’s sake: ovte Boas kre’ tperépas ovr’ aAdXov orwra. Compare 275: py) pev ey® pnt’ avTos— pyre ti’ dAAov drwmra Boov kAorbv vperepawr, 2. Lee AGAR MANCHESTER, CATVLLVS VI. 1 saa. Flaui, delicias tuas Catullo, nt sint inlepidae atque inelegantes, uelles dicere, nec tacere posses. SENTENCES Of the form sz s#t . . . esse¢ crop out at rare intervals in the literature of classical Latin, deviating so sharply from what might be called the standard forms of conditional speaking that they have commonly been regarded with more or less suspicion. Reisig, indeed, was for emending or rejecting all such cases;' and it is with considerable modesty that Rothheimer comes to the defence of the form s¢ sit... esset, Showing that it is found from Plautus onward, and making the point that the examples lend one another mutual support when brought together into a group.? Even yet editors are inclined to emend where only a slight change would be called for.* But in some cases emendation is not easy, especially in verse ; and a few occurrences of st sit . . . esse¢t have come to be accepted without qualification, among them the citation from Catullus above. Explanation of the disparity in tense use is various. For this particular passage Kiihner holds that metrical convenience is the determining factor.* On the other hand, Haase, commenting on Reisig and falling back upon the antiquated doctrine that the present subjunctive marks a condition as ‘ possible,’ elucidates thus: ‘ Sonst ist das ni sint so zu fassen, dass darin zum Scherz, gleichsam dem Flavius zu Gefallen, noch die Moglichkeit vorausgesetzt wird, dass seine deliciae nicht inlepidae sind.’® Unless memory is treacherous, someone has applied this same method of interpretation to the following case of the form si sit . . . ent: Horace, Car. III. 3. 7 sqq.: Siz fractus inlabatur orbis, impauidum fervent ruinae. The poet is here speaking of the man iustwum et tenacem propositt, and he means to say that such a one will be found undismayed under any conditions whatsoever. It is absolutely essential to the thought that the condition picture a state of affairs which is beyond any reasonable possibility. Yet, to save the theory, it is carelessly stated that Horace regards the condition as ‘ quite likely to be fulfilled.’® 1 Vorlesungen iiber lateinische Sprachwissenschaft, |wandtschaft.’ Asa matter of fact, Cicerois here § 302. 2 De enuntiatis condicionalibus Plautinis, Gét- tingen, 1874, p. 48. 3 Cf. Munro on Lucretius I. 356, footnote. ¢ ‘Hier offenbar wegen des Metrums sint statt essent,'’ Ausy. Gram, der lat. Sprache, I1., S. 924 (omitted in 2nd ed,, § 215. 3). 5 L.c., Anm. 464b. 6 The stultifying effect of unquestioning ac- ceptance of conventional generalizations is well illustrated in a different field by the treatment of the case-use with similis in the Tischer-Sorof edition of the Tusculan Disputations. There, on I. 34, the note reads: ‘sui similem. Der Gen. bei similis bez. eine vollige Ebenbildlichkeit, der Dat. eine nahe kommende Ahnlichkeit oder Ver- alluding to the story of Phidias and the shield of Minerva. Not being allowed to record his name, the artist included in the group scene on the shield a figure that suggested himself (sui similem speciem inclusit). It is altogether unlikely that Cicero means to say that he carved an exact likeness of himself—that would have been too bold (and cf. Plutarch, Pevricl. 31, atirod ria popdjv). This single note is made to do duty for the whole book; hence there is merely a cross- reference at v. 56 on the phrase hunc illi duco simillimum. Cicero is now comparing Catulus and Laelius, and, since the dative is used, the cross-reference points to a less degree of likeness, quite overlooking the fact that the adjective is in the superlative degree (simillimum). Wi of the sentenct In | clearly thought under w for the | In mixture Pla So in at Pla In of the ‘ purely differen sentenc Bu ways O circums each se Th safe.’ ' case) tk may be Fo Th come, conditic the int remitta Re esset jus on the | the spe contrall to stop 1 Amel 297 Sqq. 2 The Californi pp. 122: 3 MSS -he literature indard forms more or less id it is with | 2 Bae oint that the group. ge would be e; and a few ution, among ular passage Jn the other ted doctrine hus: ‘Sonst Flavius zu cht «nlepidae method of he means to tsoever. It te of affairs is carelessly Cicero is here 1 the shield of ord his name, scene on the elf (sui similem unlikely that ‘ved an exact ave been too I, avrov Tiva to do duty for erely a cross- hunc tli duco airing Catulus e is used, the ee of likeness, adjective is in H. C. NUTTING 87 Within recent years considerable advance has been made in the understanding of the Latin ccnditional sentence, and it seems worth while to consider anew sentences of the form sz sit . . . esset. In the first place, the history of the present ‘contrary to fact’ construction is now clearly worked out. In the time before Plautus the regular vehicle for this type of thought was the form sé sit... sit.! In the writings of Plautus the shift is well under way that ultimately made si esset . . . esset the standard method of expression for the present ‘contrary to fact’ idea.” In that period of transition it was wholly natural that there should be tense mixture—e.g., Plautus, Bacch. 635: PI. Si mihi si#,3 non polliceay? MN. Scio, dares, noui. So in an example of the form sz sit, . . esset: Plautus, Aul. 523 sq.: Compellarem ego illum, + metuam ne desinat memorare mores mulierum.* In the second place, something more has been learned of the logical applications of the ‘contrary to fact’ construction. Sentences of this category may indeed be purely speculative, announcing what would be the result if circumstances were different from what they are or have been. Thus far ‘contrary to fact’ conditional sentences are analogous to those of other classes. But they are peculiar in the frequency with which they are used as roundabout ways of suggesting thought that is in no wise conditional. This results from the circumstance that, by their very nature, a ‘contrary to fact’ protasis and apodosis are each set over against an actual fact—e.g., ‘If he were here, we should be safe.’ The first of these clauses implies ‘he is not here,’ and the second ‘ we are not safe.’ These implications are clearly felt by the hearer; and when (as often is the case) there is an obvious relation between the implications, the conditional sentence may become merely a method of znutimating the essential thought to be conveyed. For example, if A asks B for a loan, the latter may reply: ‘I would let you have it, if my remittance had come.’ Through the intimations of this sentence A learns that B’s remittance has not come, and that there is no prospect of securing the loan. And since A uses the conditional sentence as a formula of explanation or apology, the relation between the intimations is plain; the accommodation is not to be forthcoming because B’s remittance has not come.® Reference backward will show that the Plautine example of the form si sit. . . esset just cited is of this same type, being in the nature of an aside, to the audience, on the part of the unseen auditor of a long monologue on dowered wives. Ostensibly the speaker has stopped to converse with the monologist; hence at length he is contrained to offer an explanation of his delay: ‘ I should address him, did I not fear to stop his tirade on the ways of women,’ which, with its implications, conveys the 4 Cf. also the converse arrangement in Plautus, Sti. 510 (st fuisset . . . Sit). 5 Hence such a contrary to fact conditional sentence is said to exhibit the indirect causal mode ; see, further, The Latin Conditional Sen- tence, 1.¢., pp. 136 sqq. 1 American Journal of Philology, XXII., pp. 297 Sqq. 2 The Latin Conditional Sentence. University of California Publications in Classical Philology, VIII., pp. 122 sqq. 3 MSS. mihist. 88 CATVLLVS VI. 1 sgqg. information: ‘I do not (yet) address him, because I fear to stop his tirade on the ways of women,’ Cf. other such sentences of the form sz sit . . . esset: Vergil, Geor. 4. 116 sqq.: Atque equidem, extremo mz iam sub fine laborum uela tvaham et terris festinem aduertere proram, forsitan et pinguis hortos quae cura colendi ornaret canerem. Tibullus I. 8. 21 sq.: Cantus et e curru Lunam deducere temptat, et faceret, st non aera repulsa sonent. In the first of these passages the poet is explaining why he does not go on toa discussion of topics not yet treated; he leaves them out because the time is short. The other example has to do with attempts to cast a spell upon the moon, which attempts fail of their purpose (only) because the spell is broken by the beating of bronze. There is a very similar situation in another sentence which takes the form si sit... fussset: Tibullus I. 4. 63 sq.: Carmine purpurea est Nisi coma; carmina m7 sint, ex umero Pelopis non mituisset ebur. These lines are part of a tribute to the worth of poetry: poetry tells of the purple lock of Nisus, and it is because poetry exists that we know of Pelops’ ivory shoulder.’ In the light of these passages it becomes easier to analyze the lines which are cited from Catullus VI. 1 sqq. as the subject of this brief paper: Flaui, delicias tuas Catullo, nt stint inlepidae atque inelegantes, uelles dicere, nec tacere posses. Flavius is keeping very quiet as to his present love, and the poet rallies him by intimating the cause of his conduct : ‘ Flavius, it’s because your love is lacking in charm and refinement that you don’t want to tell me, and can keep your tongue still on the subject.’ As the words force upon Flavius an unwelcome explanation of his actions, they have a taunting flavour not present in the examples previously cited ; otherwise the case falls into line with the rest. Of course, Roman writers made no such formal analysis as is here attempted ; but they could and did take advantage of an undercurrent of ‘contrary to fact’ conditional speaking to develop a formula of polite regret, apology, or explanation, as illustrated above. Well attested examples of the forms sz sit. . . esset and st sit. . . fuisset are very few, and it is distinctly interesting that a common strain should be found running through at least six of them. This tends to give individuality and solidarity to the category, and militates against the assumption that sz sit... esset and sisit . . . fuisset represent a mere ragged fringe, explainable on the basis of caprice, poetic license, or metrical convenience. H. C. Nuttine. THE UNIVERSITY OF CALIFORNIA, 1 So, too, Lucretius, V. 276 sqq. A. In spontan Verse c introduc movosa, nemovosa Th So zn sol Wh Elegy i: mocker. Since eu would ft aequalem from N’ before u Nor acceptan 0 and a, tradition I propos and I fin The against ¢ and xvi. doubt up in attem] has been because « unless it on the ways ot goontoa me is short. moon, which ng of bronze. es the form tells of the y of Pelops’ es which are ullies him by s lacking in tongue still ation of his iously cited ; > attempted ; ary to fact’ explanation, . fursset are ld be found nd solidarity . esset and s of caprice, NuTTING. THE TEXT OF THE CYNTAHIA. A. I begin with some emendations founded on familiar evidence. (2) ii. 9, aspice quos summittat humus formosa colores, In each of the following verses of the stanza from wu. g to u. 14 matuval and spontaneous beauty is the subject (sponte sua, solis, indoctles, natiuts, nulla arte). Verse 9 has seemed unsatisfactory because it lacks the very point which it should introduce. All the emendations have attacked foymosa because it is repeated in w. 11: movosa, Housman (which would rather mean ‘ wayward’ or ‘crabbed’), dumosa, nemovosa, muscosa, and other even less probable adjectives have been suggested. Though 2 here happens to have spelt summiittat, not submittat, I offer as possible : aspice quos sibi mittat humus formosa coloves. So in solis formosius antvis in u. 11. ‘ For herself alone.’ (6) vii. 16. te quoque si certo puer hic concusserit arcu, (quod nolim nostros euiolasse deos)— nollim N ; e uiolasse, codd. nonnulli. Why should Propertius be wnwilling that Ponticus should fall in love? Elegy ix. (1; 9-14) shows us that in this elegy he prophesied such a fate for the mocker. The ‘violation’ has been done by Ponticus to the gods, not wice wersa. Since euiolasse is not a Latin word, it must be emended; and the simplest emendation would be @ wiolasse, i.e. est uiolasse. At III. vi. 22, where the other MSS. have aequalem nulla diceve habeve domo, N gives et qualem nullo ; and Palmer has recovered from N’s variant the certain truth: et gualem nolo dicerve habere domi. In the verse before us N has not nolim, but nollim. I propose to read: quod nulli nostros est uiolasse deos. (c) xvii. 3. nec mihi Cassiope solito uisura carinam. None among the many emendations proposed for solito has found much acceptance. The least improbable has been Housman’s (ad Mamil. I. 344) : sontem. o and a, like o and u (as above), have been frequently confused in the Propertius tradition, e.g. mauita for non ita (II. xxiv. 38); #¢ and w are at all times alike. I propose to read— nec mihi Cassiope sa/uo uisura carinam ; and I find that Lucian Mueller has anticipated me to the extent of proposing saluam. (d) xviii. 27. pro quo diuini fontes et frigida rupes et datur inculto tramite dura quies. The fact that Propertius writes montes et frigida saxa at xx. 13 is very strongly against any proposal to alter fontes here into montes ; just as the appearance at vi. 10 and xvi. 38 of an identical first half of the pentameter (quae solet ivato) throws grave doubt upon one or other line. Yet Lachmann is almost alone among the emendators in attempting to preserve fontes (and its alliteration) ; and his proposal di! niuet fontes has been rightly rejected, because it is scarcely a hardship for water to be cold, and because di/ is an intolerable interjection. What was the hardship in having water, unless it was inaccessible or unsuitable for drinking? I therefore propose to read: pro quo dumeti fontes et frigida saxa. G 90 O. L. RICHMOND Heinsius proposed dumost montes. For the construction I compare flumina stluae at xx. 7. The et of dumeti may have been written by compendium and taken for 7 (sm), in which case the number of strokes would be identical with those of diuini. (e) xix. 19. quae tu uiua mea possis sentire fauilla ! tum mihi non ullo mors sit amara loco. Even if mea fauilla could be equivalent in Latin to me faulla, what, one asks, are to be Cynthia’s feelings which will sweeten Propertius’ cupin death? Her ‘ feeling’ that the dead poet prefers her beauty to the beauties in Hades, and is only waiting for her to join them? Such feelings might comfort her, but help him not at all. Postgate referred us to Ovid, Tvistia III. 3. 84: ‘sentiet officitum maesta fauilla pium,’ What would comfort the poet would be a survival of feeling, of sympathy, 2m his own ashes after death. Compare, in a similar context again, non nihil ad uerum conscta terra sapit (II. xiii. 42). Also Lucan VII. 471: sensum post fata tuae dent, Crastine, mort. My proposal here is simply to read fossim for possis (s is following), and to punctuate before mea : quae tu uiua, mea fossim sentire fauilla! (Passerat and Carutti have used fossim before, but in rewriting the verse.) (f) xxi. 6. sic te seruato ut possint gaudere parentes : ne soror acta tuis sentiat e lacrimis .. . Whose sister ? one asks. Of course, if one reads on to the end of the epigram, it emerges that a commission is sent to the speaker’s sister, that his name is Gallus, and that the acta are the adventures of uu. 7 and 8. But w. 6 gives us no hint of all this. Even if one removes me and reads haec with Beroald, or ut (and senttet) with Postgate, one asks again, ‘whose sister?’ We have te, paventes, tuts lacrimis; the sister must, then, be the unknown soldier’s. She must be present with the parentes on his return. Yet this we know to be nonsense. I propose, therefore, to read, for the in any case intolerable ue, mz : sic fe seruato ut possint gaudere parentes : mt sovoy acta tuis sentiat e lacrimis.. . The position of sovor between mi and acta, and of both m: and sovor in relation to te and paventes, leaves no doubt that the pentameter describes a different and later scene. The corruption, after ut preceding, of m* into ne was easy enough. I may add at this point a brief remark upon the interpretation of the last couplet of the epigram. From the doubt in editors’ minds as between the quaecumque of all good MSS. and quicumque, I am not sure that any of them have understood the last four words : et quaecumque super dispersa inuenerit ossa montibus Etruscis, haec sciat esse mea. The mountains will be covered with scattered bones. Were all these the bones of Gallus? No; the emphasis is upon haec. The unknown soldier will now mark the one spot where the bones of Gallus can be found, and will report it. The haec is like the &&« of Callimachus’ Epigram xiv. 4 ; indeed, our epigram owes much to that: Kiu(ixov iv EAOns, OAtyos wovos ‘Irraxdv evpeiv kat Advpnv: adavis ov Te yap 7 yeven. Kai oi avenpodv pev epeis Eros, eurra Se AEEar Tov0’, ott Tov Keivwv Od’ Eereyw Kpitinv. The unknown soldier also fell by the way. Gallus was never found (xxii. 6-8). g) xii. 9. inuidiae fuimus: non me deus obruit ? Xi. 2I. an mihi non maior carae custodia matris ? In compar (1) ? ¥. 12, Xl. 4, et silice et § the Loe B. outline article | 1s butlt 1 with the The sct 6.6.6.6.€ itself at Viz. 6.4. No. xvii In xix. | anticipa ten, whc with the that No While r separate no differ structed us to see He has placed t! thus, 8.6 cus repe give it, t their cor this tran Each of relation - the third obscuritt concerne will be | wished t of the e conjectu: No. fall into | Nos. v, 2 mina stluae at ken for 7 (tn), utnt. one asks, are Her ‘ feeling ’ ; only waiting im not at all. fauilla pium.’ thy, a his own m conscia terra istine, mortt. ving), and to e.) he epigram, it me is Gallus, no hint of all id senttet) with ; lacrimts ; the th the parentes in relation to ‘ent and later xh, 1e last couplet vecum@ue of all ‘stood the last ese the bones will now mark The haec is much to that: xii, 6-8). THE TEXT OF THE CYNTHIA gI In each line I read nunc for non (Hertzberg proposed it in xi. 21). compare III. xviii. 7: at nunc, immsae magno cum crimine Baiae, quis deus im uestva constitit hostis aqua ? For xii. 9 (A) xiii. 33. Read with punctuation after tw: tu, uero quoniam semel es periturus amore, utere ... I accept the following familiar emendations: ii. 13, praefulgent (Baehrens) ; v. 12, sciet for solet (Itali); vili, 19, uteve for ut te (Itali); 7b. 22, ueva (Palmer); xi. 4, et modo (Itali); xiii. 24, ab Oetaeis (Scaliger) ; xv. 17, delata (Itali). xvi. 29, st silice et saxo (Voss. Lat. 81) is a reading I reported in 1910, and has been accepted by the Loeb editor. B. In my article in C.Q. for April, 1918 (XII. No. 2, pp. 59-74), I published an outline of my theory concerning Propertius’ schemes of composition. From that article I have nothing to retract. Throughout his works, the elegy, whether short or long, is built upon a scheme of stanzas balanced by the number of thety verses. I am here dealing with the Cynthia only. I analysed in 1918 the simple scheme of xiv., viz., 6.2, 6.2, 6.2. The scheme of certain others of these elegies admits of as little doubt. No. v. has 6.6.6.6.6/2. No. vi. has 4.2, 4.2, 2.4 :|| 4.2, 2.4, 4.2. No. viii. (b) uu. 27-46, declares itself at once to be a separate elegy from No. viii. (a) by possessing a distinct scheme, viz. 6.4: 4.6; No. x. has 6.4, 4.6, 6.4; No. xil, has 6.4 : 4.6; No. xiii. » 4.8, € 6.6, , 8.4 § No. xvii. » 4.6. 4||4.6.4; No. xviil., 4. 4.2, dy 2 2-4 [12.4 4, 4, 4. 1.2; No. xix. » 4. 4.2, 4. (2 |] 202 4, 2. 2.4 | 2. In xix. and in v., and, as we shall see, in ix., the closing couplet, which in its effect anticipates that of the Shakespearean sonnet, ts introduced by quare. To this list of ten, whose schemes the MSS. preserve without flaw, I can quickly add three more with the aid of Housman’s criticism. He hasseen, and certain editors have accepted, that No. i. must have lost a couplet after w. 11; he and they say actually after wu. 12. While recognizing the lacuna (the missing hexameter began e¢¢ modo) I prefer to separate tbat from ervabat, and to place the loss between uw. 11 and 4.12. It makes no difference to the present point, which is that No.1., as the MSS. give it, is con- structed thus: 8.4.6 or 8.6.4 ||6 (at wos) 6 (et uos) 8 (wos). Housman’s acumen enables us to see that the scheme was 8.6.6 (with s//e ettam beginning the third stanza) || 6.6.8. He has also seen in No. vii. that wu. 23, 24 should follow wu. 10. When they are placed there, the elegy (which in the MSS. text falls into unbalanced groups of lines thus, 8.6.10.2) is seen to have been composed as 8.8.8. || 2, with the address to Ponti- cus repeated in the first couplet of the second half. Again in No. xi. As the MSS. give it, the verses are grouped thus: 6.2; 6, 2,4; 6.4. But wu. 15, 16 do not fit into their context, and Housman places them after w. 8, where they fit very well. With this transposition the scheme becomes: 6.4 (the quatrain is uu. 7, 8, 15,16); 6.4; 6.4. Each of these three inferences by Housman has been made on grounds bearing no relation whatever to any idea of a scheme; the first two have been generally accepted, the third has been favourably regarded. If I find that besides creating sense out of obscurity they call order out of chaos, I may be forgiven for adding the three elegies concerned to those ten where the scheme already speaks for itself. (And Housman will be relieved to hear that at least one other of his proposals in this book—he wished to transfer xviii. 23, 24 to follow v. 6—is refuted by the formal perfection of the elegy as it stands, and can therefore remain in the unrestricted realms of conjecture). No. ix. can soon be added to the thirteen. As it is given in the MSS. the verses fall into groups thus: 8.6.10.8|| 2 (quave beginning the couplet, as in the conclusion of Nos. v, and xix.). Postgate bracketed uu. 23, 24, as awkward in the context. O. L. RICHMOND nullus Amor cuiquam facilis ita praebuit alas, ut non alterna presserit ille manu. But his proposal to add them to No. v. would destroy there a flawless scheme. Whereas the scheme of this elegy requires that one couplet from the third group (15-24) should be added to the second (9-14); then each group will have eight lines, My inference that uu. 23, 24 should be placed after uw. 14 is made from the conver- gence of several considerations. They have but to be seen in that position to estab- lish themselves there. i quaeso et tristis istos compone libellos et cane quod quaeuis nosse puella uelit ! nullus Amor cuiquam factlis ita praebuit alas, ut non alterna presserit ille manu. quid si non esset facilis tibi copia? ... It is the repetition of facilis in the same point of the succeeding hexameter, Housman’s ‘homoeomeson,’ even more than the w# following the e¢ in the pentameter, that has caused the omission of nullus Amor ...manu. No. ix., then, had a scheme 8.8.8.8 || 2, No. xvi. is of vital importance for my theory of the scheme, because here only we can compare the treatment of one theme by Catullus, Propertius, and Ovid. Catullus has adapted for a personal purpose a poem by Callimachus on the Door. His lxvii. is one of the expressa carmina Battiadae (lxv. 16) hardly less than Ixvi. and Ixvili (b); and like this poem of the ‘Roman Callimachus,’ Propertius, it has 48 verses. It begins with a speech of eight lines by Catullus, and ends with a speech of sixteen lines by the Door, into which have been thrust uu. 33, 34, by many seen to be alien to that context. [I can only now indicate my belief that uu. 33, 34 belong after u. 18, and that to the Door’s question (17): qui possum ? nemo quaerit nec scire laborat, Catullus answered : nos uolumus: nobis dicere ne dubita; flauos quem molli percurrit flumine Mella, Brixia Veronae mater amata meae. i.e. ‘If I am a son of Verona, I am a grandson of Brixia, and can be trusted to be both faithful and interested.’ This solution involves, of course, the inference that the Door of Caecilius is not at Verona, but at Brixia; but such an inference is natural enough from wu. 31, 32 alone.] Catullus’ poem then had 6x8 verses, like our own, and began with eight and ended with sixteen. Now let us turn to Ovid's treatment of the theme in Amores I. vi. He uses a refrain—tempora noctis eunt : excute poste sevam. It occurs at uu. 24, 32, 40, 48, 56. There are, then, stanzas of eight lines each extending from u. 17 to u. 56. In front of these stanzas are sixteen lines, or one double stanza. After u. 56 the MSS. give us, not sixteen, but eighteen lines. How- ever, the first of these couplets (uw. 57, 58) destroys the conclusive effect of the refrain in wu. 56 by continuing the thought, and in inelegant words. tempora noctis eunt: excute poste seram. aut ego iam ferroque ignique paratior ipse, quem face sustineo, tecta superba petam. It is not to be wondered at that on other grounds this couplet has been attacked ; it is a very bad one. I infer from the otherwise balanced construction of the elegy that this is a spurious couplet ; and that Ovid here used a scheme Sy 8:8, 8, 8, 8, 8: S+8, With this preface I turn to Cynthia xvi., which is likely to be nearer than either Ovid or Catullus to a Callimachean model. Like Catullus, it has 48 verses; like Ovid, it begins with 16, before the incantation proper is introduced. There is no refrain, as in Ovid (at mea : u. 24 (fi give wi. That is, the lady (How al cause, m answere break dc If this t omission two cou bottom c run, fron But The Oxf 9, 10 to f which hz reading | amantis fi hitherto. cedent of veuocatuy she make invocatio cedent of The desunt pic and to ul! asking the but great The | scheme of wless scheme. e third group ve eight lines, m the conver- ition to estab- sr, Housman’s eter, that has me 8.8.8.8 || 2, use here only is, and Ovid. on the Door. than Ixvi. and ius, it has 48 with a speech by many seen 33, 34 belong trusted to be nference that ) inference is 8 verses, like urn to Ovid’s s eunt: excute of eight lines 1 lines, or one lines. How- of the refrain attacked ; it he elegy that 8, 8, 8: 548. n either Ovid like Ovid, it oO refrain, as THE TEXT OF THE CYNTHIA 93 in Ovid; and yet, if ww. 17-24 and uu. 27-34 were two stanzas of eight lines each, wu. 34 (at mea nocturno uerba cadunt Zephyvo) would have something of that effect as following u. 24 (frigidaque Eoo me dolet aura gelu). Between these eights, however, the MSS. give uu. 25, 26: tu sola humanos numquam miserata doloris respondes tacitis mutua cardinibus. That is, they give §48:10:8 down to w. 34. At w. 35, after eight lines addressed to the lady rather than to the door, we return to the immediate enemy. Sed tu sola, mei tu maxima causa doloris, uicta meis numquam, ianua, muneribus.... (How absurd was the old punctuation of wu. 35 after met! If the door was the only cause, maxima is bathos.) At “. 35 we have, not 16, but 14 verses left; sed tu (35) is answered by at tzbt (41); but the groups are only 6+8. Is the theory of a scheme to break down? Or are uu. 25, 26 to be transposed to follow #. 36? 35 sed tu sola, mei tu maxima causa dolons, 36 uicta mezs numquam, ianua, munenbus, 25 tu sola, humanos mumquam miserata dolons, 26 respondes tacitzs mutua cardinibus ! If this transposition commends itself on familiar grounds of criticism (and surely no omission could be more easily forgiven to a scribe’s eye than that of the second of the two couplets after the first—it will have been inserted where we find it from the bottom of the page), I may infer a scheme of eights throughout our elegy. They will run, from 17-48, thus: 17-24; 27-34; 35, 36, 25, 26, 37-40; 41-48. But there is still a slight flaw in our order in the region between w. 6 and 4. 14. The Oxford editor makes wu. 11, 12, a parenthesis in brackets; Baehrens transposes g, 10 to follow 12. The problem is to account for has inter, the MSS. reading in u. 13, which has no antecedent in ww. 11 and 12. Scaliger tried to account for the has by reading has inter grauius c. d. quevelas. But no quevelae are in question yet; semper amantis fletus (4.7, 48) are contrasted with dominae uitia (47), the subject of the preface hitherto. Baehrens, then, by his transposition, would make dominae noctes (g) the ante- cedent of has inter ; and this is better, yet not satisfactory. For mec tamen illa suae veuocatuy parceve famae (11) answers nec possum defendere (g) (‘ I cannot defend her, and she makes no attempt to save herself’); and veuocatury is more pointed, if it follows the invocations of the carmina of u. 10. We are forced yet further back for the true ante- cedent of has inter; and I will give uu. 5-14 as I read them. 5 nunc ego, nocturnis potorum saucia rixis, 6 pulsata indignis saepe queror manibus; g nec possum infamis dominae defendere noctes, IO nobilis obscenis tradita carminibus. tr Nec tamen illa suae reuocatur parcere famae, 12 turpior a¢ saecli uiuere luxuria; 7 et mihi non desunt turpes pendere corollae 8 semper et exclusis signa iacere faces. 13 has inter grauibus cogor deflere querelis.... The Door is between the garlands above and the torches below. e¢ mihi non desunt picks up mec tlla veuocatuy and turpes turpior. I prefer to read at for e¢ in u. 12, and to understand there the sense of fergit out of veuocatur ; for I see no point in asking the domina to be punor saeclt luxurta, as Fonteine and others since have read, but great point in the line as I take it. The preface of sixteen verses is thus resolved into its component eights; and the scheme of the whole, I should claim, greatly adds to the force and effect of the elegy. 94 O. L. RICHMOND But I have still to rescue u. 38, upon which so many emendations have been squandered: te non ulla meae laesit petulantia linguae 38 quae solet irato dicere tota loco, ut me tam longa raucum patiare querela sollicitas triuio peruigilare moras. After years of doubt I have come by degrees to a definite conclusion. First, quae has for antecedent, not petulantia, as some have thought, but linguae. There is no ground, prima facte, for doubting the words quae solet diceve as dependent upon linguae. Second, fota cannot stand in any relation with these words, and therefore cannot be sound. The fifteenth-century scribes can only offer weyba as a variant, with turba possibly derived from this. Neither alternative helps us, if the first point is sound. Third, tvato loco is an impossible phrase, imgrato loco so frigid as to be inacceptable in so rhetorical a poem. ivato is defended by the reference to this elegy in III. xxv. 9, 10: limina iam nostris ualeant lacrimantia uerbis (=our verse 13), nec tamen irata tanua fracta manu. loco thus comes under suspicion, and the only obvious and easy emendation is zoco (Heinsius). Fourth, does the context demand anything of the word lost in ota? I submit that the consecutive ut of u. 39 is sorely in need of something to introduce it. The letters of tota then suggest ¢ata, with the common confusion of 0 and ato account for the corruption. I find that Hailer has anticipated me in proposing tanta, but do not know whether he made sense of the passage. This can only be done by putting a question mark at moras (40), and punctuating thus: te non ulla meae laesit petulantia linguae ; quae solet irato dicere tanta ioco, ut me tam longa raucum patiare querela sollicitas triuio peruigilare moras ? For tanta (uerba) cf. Plaut. Poen. 273, and also our xv. 12: imiusto multa locuta salo. For the general sense cf. our xviii. 13-16: quamuis multa tibi dolor hic meus aspera debet, non tta saeua tamen uenertt iva mea, ut tibi sim mevito semper furor . . No. iv. happens to have the same number of verses as No. xvii., which was readily seen to have the scheme 4.6.4 || 4.6.4. Here also we begin very clearly 4.6.4, for though there is an e¢ in u. 15, wu. 15, 16 do not carry on the thought from uu, 11-14, but rather hark back to the theme of wu. 1-4: uu. 17-22 are as clearly a group of six as are uu. 5-10. The last six verses involve us in several difficulties. But first as to the scheme. As the MSS. give the elegy, the verses are grouped thus: 4.6.4 ||2.6.6. If there was balance as in the fifteen elegies considered, or, if you will, as in the ten which need no consideration, then one couplet from the last group (23-28) must be transposed to precede or follow wu. 15, 16, and restore 4,6.4 to the second half of the elegy. And we have no choice; but must take, if any, uu. 25, 26, which, like uu. 15, 16, hark back to the theme of uu. 1-4, which decidedly interrupt the devotions of Cynthia (23, 24) and of Propertius (27), and which have provided even in recent editions so absurd a contradiction in vapto amore praccipue nosin (26, 27) of uu. 15 and 16 above. The passage from w, 11 to w. 16 would then run: haec sed forma mei pars est extrema furoris ; sunt maiora, quibus, Basse, perire iuuat : ingenuus color et multis decus artibus, et quae 14 gaudia sub tacita dicere ueste libet. 25 Non ullo grauius temptatur Cynthia damno 26 quam sibi cum rapto cessat amore decus ; I5 quo magis é¢ nostros contendis soluere amores, hoc magis accepta fallit uterque fide. The et of u. 1 the eleg 28 for th Cynthia stone. . maneat, nostyt ma and othe In 2 In view in tyiuto- be too h Hoefft’s No. Calypso Alphesil down to semicolc Verses < to make theory < that uu. ment of of 8, n proposa Bu to balar And uu. in 4. 33. removal belong must b UU. 25, | the wor words v here. . logical : has 8+ of 4 mu that so a lacun 1s have been ision. First, uv, There is vendent upon und therefore variant, with first point is id as to be ence to this ur verse 13), ind the only text demand - 39 Is sorely ata, with the it Hailer has sense of the as (40), and a locuta salo. meus aspera , which was learly 4.6.4, ought from e as clearly difficulties, are grouped idered, or, if om the last tore 4.6.4 to ake, if any, -h decidedly which have aecipue nostrt hen run: THE TEXT OF THE CYNTHIA 95 The loss of uu. 25, 26.at that point might be due to the repetition of decus. The et of wu. 15 would now link kindred couplets. The stanza begins the second half of the elegy by a return to the first subject. There would now remain wu, 23, 24, 27, 28 for the last quatrain. The MSS. give: 23 nullas illa suis contemnet fletibus aras 24 et quicumque sacer qualis ubique lapis. 27 praecipue nostri maneat sic semper adoro, 28 nec quicquam ex illa quod querar inueniam. Cynthia will pray for Propertius, and curse Bassus, before every altar or sacred stone. In such a context we perceive at once that w. 27 must be punctuated after maneat, sic semper adovo: the lover has his devotions too; and his prayer is praecipue nostvt maneat ! Let her continue to view Bassus as inacceptable henceforth (cf. 19), and other men as Propertius views the other girls (cf. 1, 2). In wu. 24 I agree with Housman in finding qualis intolerable: he proposed quaeret. In view of Tibullus I. i. 11,12: nam ueneror, seu stipes habet desertus in agris seu uetus in triuto florida sevta lapis, and of contemnet in our 4. 23 (which shows that no rite will be too humble), I propose to read squalet (h. e., squal;), and, as a corollary, to accept Hoefft’s alteration of e¢ into mec. The passage would run: nullas illa suas contemnet fletibus aras nec quicumque sacer squalet ubique lapis. praecipue nostri maneat, sic semper adoro, nec quicquam ex illa quod querar inueniam. No. xv. begins unmistakably with stanzas of 8.6.8 down to w. 22. (Cynthia, 1-8; Calypso, 9-14; Hypsipyle, Evadne, Alphesiboea, in that order, 15-22. The Alphesiboea couplet was copied after /aetitiae (14) because the scribe’s eye wandered down to pudicitiae at the end of the verse actually preceding it. There should be a semicolon after fratves.) But it is after u. 23 that the editors’ troubles have begun. Verses 25, 26 so seriously disturb the sense that others since Ribbeck have wished to make a new elegy commence at this point. I may therefore point out how the theory of tne scheme, if established, would affect such a proposal. It would show that wu. 1-24 are mot a complete elegy, for 8.6.8.2 would make no balanced arrange- ment of stanzas. It would show again that, as wu. 35-42 form an indivisible group of 8, no reading of the preceding 10 verses will create a balance. In fact, the proposal to divide this elegy into two would never have been made or followed. But our unmistakable 8.6.8 (down to u. 22) have here remaining only 20 verses to balance them, of which the last 8 are an indivisible group, about Cynthia's eyes. And wu. 33, 34 must have ended the preceding stanza, since hos in u. 35 refers to ocells in u. 33. Since uu. 25, 26 disturb the sense, let us omit them where they stand. Their removal creates a stanza of 8 (23, 24, 27-32). . Housman has seen that if uu. 25, 26 belong here at all, they must be placed to precede uu. 33 and 34; and that, if they must be joined, quam in 33 must be emended (with Madvig) to tam. But though uu. 25, 26 can precede uu. 33, 34 with this change, we are still left wondering what the words of Cynthia have been which she is bidden to silence. They are the same words which inspire the poet’s last injunction against blanditiae (42); but they are not here. And the theory of the scheme would solve our problem if pressed to its logical conclusion. It would say—here is an elegy in two main portions; the one has 8+6+8 verses, the other can only now be arranged as 8+4+8. But the stanza of 4 must have had 6 verses once, And our critical faculties have told us already that something which must have once appeared in our text is not here. I infer a lacuna of two verses. The passage ran (23-34): THE TEXT OF THE CYNTHIA Quarum nulla tuos potuit conuertere mores, ) tu quoqgue uti fieres nobilis historia, audax, a, nimium, nostro dolitura periclo, si quid forte tibi durius inciderit. multa prius ; uasto labentur flumina ponto, annus et inuersas duxerit ante uices, quam tua sub nostro mutetur pectore cura: sis quodcumque uoles, non aliena tamen. » (8) desine iam reuocare tuis periuria uerbis, Cynthia, et oblitos parce mouere deos, r (6) tam tibi ne uiles isti uideantur ocelli, per quos saepe mihi credita perfidia est. In No. viii. (a) the scheme would perform a similar service. The elegy begins with a clear stanza of 8; it ends with an equally clear stanza of 6 (21-26). We have already noted that viii. (b) has a different scheme, and is therefore rightly treated as a separate elegy. But between the groups of 8 and of 6 are, not 14, but I2 verses only. And here is to be found once more a familiar crux in the text. Verses 11-16 run (I have not punctuated them) : | nec tibi Tyrrhena soluatur funis harena neue inimica meas eleuet aura preces atque ego non uideam talis subsidere uentos cum tibi prouectas auferat unda ratis (15) et me defixum uacua patiatur in ora crudelem infesta saepe uocare manu. There is no subject for patiatuy in u. 15; but it should be Cynthia’s heart or herself. But note that, whether you transpose, with Scaliger, so that auva may be the subject, or read ut, so that unda be the subject (and neither sense is satisfying), you produce a stanza of 8 from u. g to wu. 16; and the poem is then grouped as 8+8+4+6. Admit the scheme as aid, and it tells you with certainty that a couplet has been lost before u. 15. ‘The second stanza should be of 6 verses, like wu. 21-26; the atque (13) suitably reinforces the last of the poet’s prayers, which end at vatis! Then followed a stanza of 8, corresponding to wu. 1-8; the first of its couplets is lost, the second was linked to it by the e¢ of wu. 15. Scheme: 8.6 | 8.6. I have covered in this survey all the elegies of the Cynthia except Nos. 2, 3, 20, and 22. What I have to say about three of these would introduce too late in this paper a factor which I discussed at length in C.Q. for April, 1918—the pagination of the uncial MS. which lies behind our archetype. Also it may be well not to anticipate too many details of my edition of the text, upon which I am now again continuously at work. I shall be greatly obliged to any scholar who will work out with a text before him the schemes and patterns which I have indicated, but not elaborated, here, and will inform me of any negative criticism of my results which study suggests. O. L. RicumMonp. EDINBURGH. ) peTpyoas aieu ETL K TpOTEpov It i: fact, unle principle and the which th these co: these las been not text at it about fre to have « tion. B treaty be these tw Atheniar restore e between appeals against | argues t Euryme Egyptia would n no chan; of Magn lived ; tl bonds of was able Lampsa tribute t Of aware. 480 B.C.- down to tion, for could doa remarka The the lang emendat assessed the achi egy begins '-26). We ore rightly not 14, but 1 the text. > heart or ya may be atisfying), rouped as it a couplet 11-26 ; the s/ Then s lost, the - 2, 3, 20, e in this ination of 1 not to OW again work out , but not ts which [MOND, TWO NOTES ON HERODOTVS. I. ’Apradpévns .. . kal Tas xwpas ofewv perpioas kata Taparadyyas. . . KaTa dy TOUTOUS petpioas popous erage Exdorouw, of kata xwdpnv SiateAéovor ExovTes Ex TOVTOY TOU Xpovov aiei Ere kat és éue ws erdyOnoav é€ ’Aptadpéveos + érdxOnoav Se oxeddv Kata TavTa Ta Kal mporepov efxov.—VI1. 42. Ir is not, in general, permissible to alter a text to suit historical or geographical fact, unless there is something in the language itself which also demands a change: a principle which justifies, for example, the insertion of a negative in Thuc. VIII. 101.1 and the obelizing - «repos in Ar, Const. Ath. 24.3. But there are also cases in which the error is certain and the author most unlikely to have committed it; if both these conditions are present, we may suspect the text. It is generally agreed that these last two couditions are present in this passage of Herodotus; but it has not been noticed (I believe) that the first is present also. Blakesley indeed accepted the text at its face value, and considered it a triumphant proof that the boasts of Athens about freeing the Greeks of Asia from the barbarian were all false—a fact that seems to have escaped her contemporary critics; and others have toyed with this explana- tion. But if the direct statement of Thucydides (VIII. 5. 5) and the terms of the treaty between Persia and Sparta (VIII. 18) are not sufficient to refute it, consider these two facts: (i.) Erythrae had an Athenian garrison, from about 460, and the Athenian commander had some control even over her internal affairs; she could not restore exiles hostile to Athens: did she pay tribute to Persia? (ii.) In the quarrel between Miletus and Samos in 441, it is Miletus, the city on the mainland, that appeals to Athens; and Pissuthnes and the Samian oligarchs prepare to march against her. Why, if the Persians were strong enough to exact tribute? Macan argues that many new cities became tributary to Athens for the first time after Eurymedon, and many were lost again and became tributary to Persia after the Egyptian disaster. But neither of these changes affected Ionia; and if they had, it would make Herodotus’ statement even more untrue, for he asserts that there was no change. Macan also adduces Thuc. I. 138. 5, the gift by Xerxes to Themistocles of Magnesia, Myus, and Lampsacus. But note that it was only in Magnesia that he lived ; there was his monument; and thence only came any revenue to him. The bonds of the Delian League were not all of them very tightly drawn: Themistocles was able to pass through Ephesus on his way to Persia; Anaxagoras could retire to Lampsacus when driven from Athens; but the cities of Asia Minor did not pay tribute to the King. Of this Herodotus, Ionian, contemporary, much-travelled, must have been aware. How and Wells think he is referring to the time of his childhood, circ. 480 B.c.—‘ just as a Bengalee might now write “ the rule of the John Company lasted down to my time,”’ without noticing that their example contradicts their explana- tion, for they use not the present tense, like Herodotus, but the past (just as Herodotus could do on other occasions, IV. 124, V. 45. 2, 89. 3). Besides what would there be remarkable in a systam of taxation which lasted from 493 to 480? There is then something wrong with the text, and fortunately there is that in the language which suggests the error and supplies as well the first condition of emendation. At the end of the passage here quoted we are told that they were assessed in practically the same manner as they had been before. Wherein, then, lay the achievement of Artaphernes? We read that he used a novel method of assess- 98 TWO NOTES ON HERODOTVS ment, xara wapacdyyas ; then that it was only the old system over again.!_ But read éraxOnoav yap oxédov ard taita ra Kal mpdrepov efyoy, and you have an intelligible sentence. If the Athenians took over in Ionia the system of Artaphernes,* that was something worth recording, something that the satrap could be proud of, as of the pacification of the cities. It has been suggested before that Herodotus was referring to the tribute to Athens, but the text as it stands will not support this, and my argument is that the language of the text itself demands some change. II. Atywwnréwv S€ ci rayxeés Eravarrdvtos ode Tov Sipov dua Nixodpéum érexpdrynoay, kal érerté ogeas xetpwodpevor eLipyov drodéovres: dad rovrov 8 Kal dyos odu éyévero, 7d éexOicar Gar ob oioi te éyévovto ériynxavapevor, GAN’ EPOnoav éxmeadvres mpdTEepov ék THS vnocov 1) ode iAeov yever Oar Tv Géov.—VI. gt. Everyone, so far as I know, supposes this to be a reference to the evacuation of Aegina by Athens in 431 B.c. But I find it difficult to believe that Herodotus could have referred to so notorious and recent an event so vaguely, without even #7’ ‘A@nvaiwv or xpdvy toetepov. Nor does Thucydides in his narrative of the expulsion say anything about an dyos, an excuse that the Athenians would gladly have made use of, then and earlier. The subject of éxrerdvres must be not the Aeginetans in general, but ot wayées (who are also o¢x in ayos ou éyévero and below), and the refer- ence will be to some now forgotten revolution in Aegina in which the nobles, like the Alcmaeonidae in Athens, were expelled. This might, of course, have happened when Athens conquered the island in 458, but even so I should have expected tz’ ’AOnvaiwv after éxzerovres (the more especially as this dyos was incurred during the early struggle with Athens). I hope no one will suggest that, if I must correct Herodotus, I should leave the first of these passages as it stands and insert ia’ ’A@nvaiwv in the second. A. W. GomME. GLASGOW. 1 Busolt and others translate ‘The Ionians for xara raird. paid as much as they did before,’ though assessed 2 We may compare perhaps Plut. Arist. 24, after a different method ; but this is not what mpocératay airo ywpay te xal rods mpogddous Herodotus says; we should want at least rocaira émioxeWduevor dpicat, xTr. POLYBIVS AND A LITERARY COMMONPLACE. TuHIs paper is a contribution to the question of how far Polybius fulfilled that part of an historian’s duty which consists of acquiring information; it gives an instance, small in itself no doubt, where he definitely neglected to obtain good information which lay to his hand, and preferred to repeat a commonplace untruth of the literary hacks. Polybius V. 84 (battle of Raphia): Ptolemy’s African elephants were defeated by Antiochus’ Indian elephants: rjv yap dopijv Kat dwoviv ov peévovor (the Africans), GAAG KatarerAnypevor Td péyeOos Kai tiv Sivapev, ds y’ epol Soxei, fevyovow evews ef dzroarnpatos Tovs "Ivdtxovs €Xéhavtas* 5 Kal Tore ovveBy yeverOor. I am not considering why the African elephants were defeated (they were heavily outnumbered), or how Polybius knew that they were in the habit of running away from Indians when the two species had never before (or I believe since) met in battle; all I want to notice here is tl Polybius to himse! Livy 37, reserve ; Africi re animorur modern \ in disput recorded known I Firs Ptolemie Both sug (1) we may But we Geds veo 20 mina average authorit' possible recordec Helleni: would t average a Germ 9 kgs.— Lane, < confirm tusks, t away h: (Athen. least, a: (2) elephar for trai war el Raphia the Pt of the to exp! a know Philote influen anyon¢ may h years on Ar therefc 1 Th. bungen But read ov, and you 2 system of atrap could before that nds will not nands some > TTEK PATNO AY, > , | €yYEveToO, TO repov €K TIS acuation of dotus could it even tz’ e expulsion have made yinetans in | the refer- es, like the ened when ””* AOnvaiwv the early | leave the GOMME. t. Arist. 24, $s mporddous ,E. filled that gives an ain good intruth of defeated Africans), evbéws ef nsidering ), or how when the to notice W. W. TARN 99 here is the statement that the Indian elephants were Jargey. It is not a case of Polybius merely copying out his authority for Raphia ; he appropriates the statement to himself—ds y’ €uot doxet—and repeats it even more vigorously in another place : Livy 37, 39, 13 (battle of Magnesia), the Romans put their sixteen elephants in reserve ; for, besides that Antiochus had fifty-four, ‘ne pari quidem numero Indicis Africi resistunt, siue quia magnitudine (longe enim illi praestant) siue robore animorum uincuntur.’ Regard for my own species forbids me to enumerate the modern writers who have repeated Polybius’ error as though it were a fact. It is not in dispute that the African elephant is often the larger of the two; and cases are recorded of the African being nearly a foot taller at the shoulder than the tallest known Indian. First, is there any chance of Polybius’ statement being true? That is, were the Ptolemies’ elephants either some smaller species now extinct or immature animals? Both suggestions have been made. (1) No trace of such a smaller species in East Africa has been found, for I think we may neglect the distant and somewhat enigmatical pigmy elephant of the Congo. But we have now, in fact, exact evidence about the Ptolemies’ elephants. Ptolemy Geds veos Aidvvoos gave to the temple at Didyma 34 tusks, weighing 24 talents 20 minas.! Wiegand, using the Ptolemaic talent, makes the total 482 kgs., and the average per tusk 14°4 kgs. (=31°7 lbs.). But, looking at the way the temple authorities at Delos regularly weighed their anathemata themselves, it is quite possible that the temple authorities at Didyma weighed the tusks themselves and recorded their own weighing. If they used the Attic talent, usual in Asia in Hellenistic times, of 26°16 kgs. (the old Milesian talent was far larger), the tusks would total 636°5 kgs., and would average 18'1 kgs. (= 39°9 lbs.). The tusks then averaged per pair either over 63 lbs. or just under 80 lbs. Wiegand cites from a German work the average weight of African tusks now coming to market as 9 kgs.—say 39 lbs. the pair; and Messrs. Lewis and Peat, ivory merchants, Mincing Lane, of whom I enquired, kindly informed me that they could ‘more or less confirm’ this figure. (Exceptional tusks do not come into question.) Ptolemy's tusks, then, were obviously a very fine average; and naturally he would not give away his best or exceptional tusks, seeing that they would be exhibited on occasion (Athen. 201 A). The African elephants of the Ptolemies were then, at the very least, as good as those known to-day. (2) Of course the elephants would be caught and trained young. The African elephants at Apia, in the Belgian Congo, which plough and draw waggons, are caught for training at about twelve to fourteen years (estimated).?, But, assuming that a war elephant would have to be forty, there was more than enough time before Raphia in 217 for the animals used there to have matured, for the commencement of the Ptolemies’ elephant hunting can be dated early in Philadelphus’ reign. The first of the long row of officials sent to the Troglodyte coast was Satyrus, as he was sent to explore. Ptolemy II. sent him ézi rijv dcepedvnow rhs tov eAepavrwv Ojpas (already a known thing) kai rijs TpwyAodutixijs (Strabo 769). He called the station he founded Philotera, after Ptolemy’s sister. Now after the marriage of Arsinoe II. in 276/5 her influence was so great that Satyrus could not have called his post after her sister or anyone but herself. Her marriage may even be earlier ; and, in any case, her influence may have antedated the marriage—probably, therefore, Satyrus’ mission was some years before 276. After Arsinoe’s death in 270 it cannot be, as Callimachus’ poem on Arsinoe’s death shows that Philotera predeceased her sister. Ptolemy II., therefore, whom tradition represents as the real amateur of war elephants in the 1 Th. Wiegand, Siebenter Bericht uber die Ausgra- number). bungen in Milet und Didyma, 1911, p. 50 (no 2 The Times, July 12, 1924. S219834 100 POLYBIVS AND A LITERARY COMMONPLACE dynasty,’ began to hunt, and therefore to train, them by 276 at ieast, and possibly some years earlier. This is confirmed by the fact that somewhere between 269 and 250 he threw enough African ivory on the Aegean market to break the monopoly previously enjoyed by Indian ivory and cause a tremendous slump in the price.” Polybius’ statement, then, being untrue, where does it come from ? The farthest back I have been able to trace it is a passage in Ctesias. Diodorus quotes the passage twice (2, 16, 4 and 2, 35, 4). The exactly parallel language shows a common source, which indeed extends (as comparison shows) to the generality of these two extracts about India; and 2, 16, being part of the Semi- ramis story, shows that that common source is Ctesias.2 In 2, 16, 4 the Indian elephants rais rot cdparos popars ToAd mpoexovor Tov év TH AtBin yevopévwv; but the other passage is fuller: not only rais popas ra Onpia tadru word mpoéyer TOV KaTa THY AcBunv yevopévwv, but they are also peyiorovs—i.e. larger than the Africans—and Ctesias gives the reason, viz. because India produces so much food. Ctesias then thought the African elephants lived where food was scarce—i.e. in the desert. This shows, if proof be needed, that he knew nothing about the African elephant, and naturally so, for his Persian informants would know nothing either. Ctesias is merely lying in his usual style for the honour and glory of his Persian employers, who did own Indian elephants. The story is, as one would expect, repeated by Onesicritus (Strabo 705), and for the same reason: the Indian elephants are larger than the African xai éppwpeverrépovs (Ctesias’ pops). He knew nothing of African elephants either, but his hero Alexander owned Indians. Curtius 8, 9, 17 repeats the story, ultimately no doubt from Onesicritus, but through a medium which knew of the Ptolemies’ elephant training: ‘elephantorum maior est uis quam quos in Africa domitant, et uiribus magnitudo respondet’; and Pliny 8, 9 says: Indicum Africi pauent, nam et maior Indicis magnitudo est.’ Pliny’s source might go back to Polybius ; but I believe there is no trace of Curtius knowing him. Strabo 705 mentions that ‘others’ beside Onesicritus said the same. It was then a literary commonplace originating with Ctesias, and repeated by Onesicritus and who knows what other fabulists; but no statement could well have a worse origin than those two. Polybius may have met it in a dozen places; he had read many writers who will never fall to our lot. But even if he could get no better information about Raphia itself, many a Roman officer could have told him the facts about the size of African elephants and the reason they were kept in reserve at Magnesia; probably a simple question to his friend Scipio would have sufficed. Instead, he has perpetuated a literary commonplace, untrue, and drawn ultimately from two of the least trust- worthy writers Greece ever produced. One is almost sorry one noticed the fact. W. W. Tarn. 1 Diod. 3, 36, 3; Strabo 769, 770; App. 2 The relevant prices at Delos are: in 276, Prooim. 10; O.G.J. 54, 1. 12, where mp@ra for 8 dr. the mina; in 269, 8 dr, 2 ob.; in 250, mparos is certain. That papyrological evidence 3dr. 3 ob. J.G. XI. ii. 163, 1.7; 203 A, 1. 71; only begins later dues not of course invalidate 287A, 1. 118. this. 3 Jacoby, Ktesias in P.W. I H: above pi translati give the ‘design. quality, : I do no a perso! suggests striding shall pe the phre may adc lofty m persiste ule he 8 spoken Th exampl uiuendi Velleiu: mus,’ a proposi where warned ments | is, I thi ‘Profe: * course bad, br But otl adopte up a cc of cert accord: is, I su than a viction C. in De ¢ action’ grauit 1 Ste sito’ as right, a are twc though ind possibly een 269 and € monopoly - price.? Diodorus el language ws) to the f the Semi- the Indian wv; but the OV KaTa THY ‘icans—and ‘tesias then sert. This phant, and Ctesias is employers, epeated by are larger of African I7 repeats hich knew n quos in : Indicum t go back Strabo 705 a literary tho knows igin than s who will ut Raphia of African 7 a simple etuated a ast trust- fact. TARN. re; in 276, »; In 250, 3A, 1. 71; ‘TENAX PROPOSITI.’ I HAVE never read the two great stanzas of Odes III. 3 without a feeling that = above phrase was rather inadequate, according to what I suppose to be ree gi translation. I base the word ‘accepted’ on Forcellini, and Lewis merci ort, w ° give the reference under the head of propositum, ‘purpose,’ ‘ intention, oe ‘design.’ But the capacity of sticking to some particular purpose is not a —— noble quality, and if we take the phrase in the wider sense of general tenacity of purpose, I do not feel that we get the ‘iustus’ of the ode, though we my perhaps get a person who is likely to succeed in life. Taken in this way, ‘tenax ne somewe suggests (as much as anything) the portrait of Henry Vill, with his — se se striding legs, dominating the English Church and the High Table o vm y: . shall perhaps be told that the context shows the meaning. It may no dou t ge the phrase from being misleading, but it does not make it worthy of its setting. may add that in the other two places, where I know of its occurring, it has = — lofty meaning. In Ovid Met. X. 415, ‘propositi tenax is used to descri eo e persistent efforts of Myrrha’s nurse to find out what is the matter. In Saints a XI. 1, 90, in a discourse on mild equivalents, we are told that the ‘ pertinax’ may be ‘ultra modum propositi tenax.’ "a econ mer ei ero I think, to be transferred to another set of examples, which Forcellini gives under the heading ‘ refertur etiam ad institutum uiuendi,’ a way or course of life, ‘a profession, school, sect. For this he quotes Velleius II. 2, ‘uir uita innocentissimus, ingenio florentissimus, proposito wapremeed mus,’ and (with other examples from Phaedrus, Seneca and Juvenal) med graue propositum ’ of Hippolytus, Iuu, X.625. He might have added Quintilian X he 2. where the professed philosophers (‘philosophiam ex professo nt or A warned that people do not like fine oratory from them. Such inapeerreen om . ish- ments are ‘tali proposito diuersa.’ As to F orcellini’s renderings, school - ) — is, I think, wrong. The ‘propositum’ is not identified with the body oe olds it. ‘Profession’ in the religious sense might suit the Quintilian passage. Way,’ or ‘course of life,’ grasps one side of the meaning. The ‘propositum ’ may me good or bad, but in either case it is for life, or some side of life, not for single meester But otherwise the rendering is inadequate. The ‘ propositum Is always sconce adopted, and is a great deal more than the bundle of habits which my make up a course of life. In the case of the philosopher it implies the deliberate ~ option of certain principles of life and the deliberate choice in each action of w . is ws accordance with those principles, If Horace is voicing the Stoics, as ees e is, I suggest that ‘ true to his creed and rule of life would give the ee etter than any single word. The ‘propositum’ is a creed in so far as it is a body of con- victions, and a rule of life in so far as it must be expressed in a body of actions. Cicero does not perhaps use the single word in this sense, but he comes eK it in De Off. 1. 112. After extolling ‘ aequabilitas cum uniuersae uitae ee singu — actionum,’ he quotes the case of Cato: ‘ Catoni cum incredibilem tri uisset na “oe grauitatem, eamque ipse perpetua constantia roborauisset, semperque in proposito is * - nere consilium’ it seems improbable. na spe man Oe aes mea te i ch “ po rather surprised that po — is a right, and that propositum and susceptum consilium, quoted by commentators on Horac . 3, 1-8. are two separate attempts to give mpoalpeois, but There are some Se though though Cicero does not elsewhere use ‘pro- ‘uultus tyranni ' is given a different turn. 102 ‘TENAX PROPOSITI’ susceptoque consilio permansisset, moriendum potius quam tyranni uultus aspiciendus fuit.’ Here the ‘ propositum susceptumque consilium’ is evidently not a single intention, but the principles which governed Cato’s life, and is the equivalent of the propositum ’ of the writers above quoted. I take it that of ‘propositum,’ in this sense as well as in the ordinary sense, mpoaiperis is the Greek equivalent. In Epictetus, indeed, in passages which other- wise remind us of this ode, the rpoaipeous, which Zeus cannot conquer and of which ov Ayors ov TYpavvos yiverat, seems to be the will which itself rpoa:peirax. On the other hand, in Lucian De merc. cond. 4, we have philosophers described as éréc-ou orovSatorépav THY Tpoatperiy mpociAovro, where, as in Quintilian XI. 1, the word might be translated ‘profession,’ and implies a course of life based on definite principles. F. H. Cotson. CAMBRIDGE. GLEANINGS FROM GLOSSARIES AND SCHOLIA. I. LVCILIVS 1134-6. My hope (Class. Rev. 34, 105) of an edition of the quotations in the Lrber Glossarum has at last been realized in Professor Mountford’s excellent Quotations from Classical Authors in Medieval Latin Glossaries (= Cornell Studies in Classical Philology, Vol. XX1.), New York and London (Longmans, Green and Co.), 1925. For the Lucilius fragment (known to us only from this glossary) oa the famous ball-player, Coeliu’ conlusor Galloni scurra trigonum tCum ludet solus ludet et eludet, t he provides a various reading of the last three words /udet ecce ludet. But this variant gives no help. Nor does Professor Housman’s (not very happy) conjecture (C/ass. Quart. I. 158) sctus for solus. Surely, if so awkward a sentence as cum ludet sciw’ ludet et eludet had been the traditional reading, the archpriest of the emendator-sect would have been the first to alter it. Marx in his edition of Lucilius explains ludet et eludet as ‘solus pilam iaciet simul et iactam ab alio repellet sinistra.’ Does he not mean exctpret simistva ? My purpose in writing this note is to call attention to Donatus’ statement in the Terence-commentary (on Eun. 55) that eluderve was a technical term for winning a game: ‘eludere proprie gladiatorum est cum uicerint . . . Eludere est finem ludo imponere.” This sheds a ray of light. Is there not a contrast between conlusoy and solus ludet? Mlle. Lenglen in a Doubles ‘cum ludit, sola (nam apparet ludere sola) ludit et uincit.’ Or, giving three facets to the verb /udo, we may say of this player ‘conludit, sola ludit, et eludit’ (i.e. wins the match), Professor Mountford hints that the gloss may come from a Virgil-scholium (of Donatus ?). The and I su} the canal or a SWO! Fest definitior form: The is a frau those of fragment marks w largely f which is ? Not makers monaste: miserab] Rec was me recognit edition) such a: is merel my prev Nu The Be And so insertio1 dicas mt All vitiated The wa Bu of these aspiciendus ot a single alent of the inary sense, which other- id of which On the other “T OVOGLOT Epa e translated CoLson. ILIA. v Glossarvum om Classical Vol. XXL), the famous his variant ‘ure (C/ass. t sciw’ ludet sect would et et eludet ;not mean lent in the winning a inem ludo nlusoy and dere sola) his player -scholium W. M. LINDSAY II. TAMA ‘VARICOSE VEIN.’ The word is known only from Lucilius’ line (1195): Inguen ne existat, papulae, tama, ne boa noxit, and I suppose those conjectural emendators, who thrive in our foggy climate and by the canals of Holland, will change it to ¢wma, since it seems to mean a swollen vein or a swollen ankle. Festus’ epitome of Verrius Flaccus’ ‘ Dictionary of Republican Latin’ gave a definition of it, and Festus’ definition has been shortened by Paul the Deacon to this form: Tama dicitur cum labore uiae sanguis in crura descendit et tumorem facit. The Schedae Pomponi Laeti offer these words as the words of Festus. But this isa fraud. The Italian Humanist could not decipher Festus’ words and substituted those of Paulus. I have just found a clue to the actual sentence of Festus from a fragment of a Bavarian glossary, compiled in Charlemagne’s time. This compilation marks with the letter LZ the glosses taken from Abolita (and A bolita, we know, drew largely from Festus). In the TRA-section appears (marked with L) this item: Trama: tumor crure et duritia et indentione ruorum ex lassitudine, which is, I suppose: T.: tumor in crure et duritia et intentio neruorum ex lassitudine. Not that we may assume these to be the ‘ipsa uerba’ of Festus. Glossary- makers had no hesitation in recasting and simplifying an interpretation for their monastery-pupils. Still, they supply new evidence to the text. And they show how miserably defective our two MSS. of Abolita are. III. ON DONATVS’ TERENCE-COMMENTARY. Recently in this journal (XIX. 28) the worthlessness of the Donatus-lemmas was mentioned; and my reason for returning to this subject is partly a_ fuller recognition of the mischief they have made (even after Wessner’s small Teubner edition) for students of Terence. No statement that ‘ Donatus approved such and such a reading’ should be accepted without consulting Wessner. If the reading is merely in the lemma, the statement should be dismissed. The example given in my previous article was (Eun. 1064 quod dicas mtht) : NIHIL EST QVOD DICAS MIHI. Aut deest ‘ non,’ ut sit non est quod dicas mihi, etc. The Bembinus has: Si te in platea offendero hac post umquam, quod dicas mihi ‘ Alium quaerebam, iter hac habui’: periisti. And so, we may suppose, had Donatus’ MS. The unsuitable lemma is a later insertion from some Carolingian MS. of Terence. These MSS. have mthil est quod dicas mtht. All this ought to be common knowledge, but I now realize how the lack of it has vitiated nearly every recent discussion of the Ueberlieferungsgeschichte of Terence. The warning therefore must be repeated. But there is another danger for readers of the Commentary. The inconsistency of these adventitious lemmas with the scholium would occasion Carolingian inter- 104 W. M. LINDSAY polation. At Andry. 536 (where the Bembinus is not in evidence) the minuscule MSS. have: Ausculta paucis, et quid ego te uelim et tu quod quaeris scies. (To improve the metre editors delete the first e¢.) Our (miserably reduced) MSS. of the Commentary show merely: AVSCVLTA Pavcis. Et ‘ paucis’ et ‘pauca’ legitur. But we get a fuller glimpse at what Donatus actually wrote when we read Priscian’s two notes (II. 281 and 320) on the construction of ausculto (Dat., Acc.). This was mentioned in my previous article, but I now wish to add the hint that Donatus may never have written a single word about the variant paucis. (He defended pauca.) Suppose a Carolingian monastery-teacher (praeceptor or magister) engaged in cutting down the Commentary to handy size. He finds in the monastery MS. of Terence at this line ausculta paucis, and makes this the lemma. He finds in Donatus’ scholium a defence of ausculta pauca. Would he not dispose of the whole thing by writing thus? (AvscvLTa Pavcis. Et ‘ paucis’ et ‘ pauca’ legitur.) Not everyone, I admit, will answer ‘ Yes.’ For paucis (Abl.) is offered with verbs of hearing by the MSS. of Terence (for Plautus,! only with verbs of speaking): Eun. 1067 prius audite paucis. Hec. 510 ~=audi paucis. And the first e¢ in the phrase et paucis et pauca legituy weakens my theory. (So the great Latin Thesaurus’ reference to Ad. 806 need not have been questioned.) Here is a more convincing example: Phorm. 469 nequid .. . poteretur mali. So the Bembinus. The minuscule MSS. read fatevetur. The Commentary (as we have it) is: PATERETVR MALI. Legitur et ‘ potiretur,’ quia potiri tov péowv fuit. Plautus ‘hostium potitus est.’ What right have we to affirm that Donatus recognized the variant pateretur ? Again, let the fancy create a Carolingian epitomator of the Commentary. Would not his attempt to join the lemma (provided by his monastery’s text of Terence), PATERETVR, with Donatus’ scholium (on fotivt) necessitate some such connecting- phrase ? What, then, is an editor of Terence to do? A glance through the Commentary is enough to show the prevalence of this type of note: legitur et, etc. Is there here any real evidence of a variance in Donatus’ day? (Cf. Arens’ Muenster- dissertation.) Take another example, Phorm. 82, where all our MSS. of Terence have: hanc amare coepit perdite, Here Arruntius Celsus attested (apparently) avdeve (for amare), as we learn from Charisius (213, 16 K.; or rather 276, 7 B., for we must all now use Barwick’s small Teubner text): Perdite pro ualde, ‘eam amare (leg. ardere ?) coepit perdite’; nam ita Arruntius Celsus et addit: ‘ antiqui enim dicebant ardere pro amare.’ The monastery Terence here showed (a mere mistake) perditus. In the epitomized Commentary accordingly this hash is served up: Aul, 1; Capt. 53; Men. 252, 779; Pers. 599; Rud. 120, 650 ?, 1102; Trin. 4; Truc, 864. AM. Who wi amare pe: In» cviticus 1 their ev existenc minuscu writers, F in let me at form of an ear tc demolisl The saec centurie: ascribed note on tenth (? reader ’) appulrt, Pro ‘Is Don least the Mer I of except F Mer Pro Scholia Virgil ’) are a pl Abstrus: line. B Mer I ol like a by (galgulus minuscule 1) MSS. of | Priscian’s This was natus may led pauca.) in cutting Terence at ’ scholium by writing fered with speaking) : . (So the ) ary (as we Plautus pateretur ? r Would Terence), onnecting- mmentary there here Muenster- 2arn from ck’s small Arruntius :pitomized ;, 864. GLEANINGS FROM GLOSSARIES AND SCHOLIA 105 AMARE COEPIT PERDITVS. Legitur et ‘perdite.’ Et ‘amare perdite’ utpote puellulam et citharistriam. Who will go bail that Donatus did not write something like this? Ardeve perdtte] amare pevdite, utpote puellulam et citharistriam In the new Oxford Terence I do not dare to exclude wholly from the apparatus cviticus these doubtful Donatus-variants, but I add to them a query-sign to disable their evidence. And if their evidence goes, what remains of the evidence for the existence in Donatus’ time of a Terence-text (more or less) like the text of our minuscule MSS.? ‘Like the text of the Calliopian edition’ say the dissertation- writers, who forget that much in our minuscule text is of mere Carolingian origin. Finally, since an important article in Evanos (of 1912) may have escaped notice, let me add that there is not now any evidence that our extant (reduced and composite) form of the Commentary is pre-Carolingian. Wessner, in an unlucky moment, lent an ear to arguments based on scholia in the Bembinus. These arguments have been demolished by the great Latin scholar of Sweden, Professor Léfstedt (Evanos XII. 43). The saec. VJ..in Wessner’s Preface (pp. xxxiii, xxxvii, xlv, etc.) is at least two centuries too early. Our earliest MS. of the Commentary (I have not seen it yet) is ascribed to the eleventh century and the Fleury scriptorium. Therefore the curious note on Andy. prol. 1, im secunda lectione apulit fuit, may mean merely that in a tenth (?)-century text of Terence (or of the Commentary) a corrector (‘second reader’) had written either dpulit (i.e. aspulit) or else attulit as his emendation of appulrt, IV. DONATVS’ SCHOLIVM ON GEO. IV. 14. Professor Rand’s article in this journal some years ago (C.Q. X. 158 [1916], ‘Is Donatus’ commentary on Virgil lost ?’) allows us to label by Donatus’ name at least the first half of the extant scholium (ed. Thilo, 1887) : Meropes aliaeque uolucres] Meropes rustici tbarbarost appellant ... sunt autem uirides earum pennae, et uocantur apiastrae quia apes comedunt. I obelize barbaros because there is no other record of birds called by this name, except Probus’ scholium (Donatus’ source) on the line: Meropes dicuntur aues quas in Italia uocant barbaros, quarum uox quia multi- plex capit partitionem uocis, meropes appellantur Graece, etc. Professor H. J. Thomson has told us to look for Donatus-material in the Berne Scholia (Journ. Phil. XXXV. 257, ‘A new Supplement to the Berne Scholia of Virgil ’) and in the glosses culled from a similar collection of scholia. These glosses are a prominent part of the Liber Glossarum and come into it from the (full) Abstrusa Glossary. No gloss, so far as I know, conserves Donatus’ remarks on this line. But here is the Berne scholium (or rather patchwork of previous scholia) : Meropes] {Galbeoliy, ut putat Tranquillus. Hae genitores suos recondunt iam senes et alere dicuntur, in similitudinem ripariae auis, quae in specu ripae nidificat, ut in libro X ostenditur. Meropes] Haec merops. Merops rex qui ob scelera ira deorum in auem sui nominis conuersus est. Meropes] Virides; et uocantur apiastrae qui comedunt apes: Gaudentius dicit. I obelize galbcolt because this form is not found elsewhere. It looks, however, like a byform (or wrong form) of one of the two names of the golden oriole, galbus (galgulus was the other), just as two families of the MSS. of Martial offer as the title- H 106 GLEANINGS FROM GLOSSARIES AND SCHOLIA heading of XIII. 68 galbuli, while the third shows correctly galbt. A common spelling of galbus was galuus, as in the bilingual gloss: Galuus (corrected from Galucts) : xAwporrpovOiov. And from galbus or galuus was derived the adjective for ‘yellow,’ galbinus (Fr. jaune). Suetonius’ Prata (his great encyclopaedic work) is presumably referred to, though we cannot be quite sure of in libyvo X, since the word Plim may have dropped out. For Pliny in Bk. X (paragraph 99) of the Naturalis Historia says: Merops uocatur genitores suos reconditos pascens, pallido intus colore pinnarum, superne cyaneo, priore subrutilo: nidificat in specu sex pedum defossa altitudine. Here apparently is the fountain-head. Suetonius identified the merops with another bird, the golden oriole, I suspect; though the golden oriole is quite unlike the bee-eater. Donatus—like Mr. Pott’s reporter—combined the information. Professor Wessner and Dr. Philipp have taught us that Isidore took material for his Etymologiae from Donatus’ commentary on Virgil. Straight from this scholium of Donatus comes Etym. XII. vil. 34: Meropes, eosdemque et jtgaulost, qui parentes suos recondere atque alere dicuntur. I obelize gaulos because gaulus is known as the name of a tubby ship, but not as the name of a bird, and I conjecture the emendation ga/wos. But whether the corruption of galuos to gaulos was made by the scribe of Isidore’s Donatus exemplar, or by Bishop Braulio the editor of Isidore’s encyclopaedic work, or by some later scribe, who knows and who cares? It is enough that we have another witness to the name of the bird identified with the mevyops by Donatus (who borrowed the information from Suetonius). The gaulus attested by Isidore and the galbeolus attested by the Berne scholia must, each of them, contain a portion of the truth. And is not the whole truth that Donatus wrote galbus or galuus ? Now I hope that someone with more courage will say: The barvbavos of Servius Danielis and of Probus on Georg. IV. 14 is a corruption of galbos. I myself do not dare to say that until we know more about these pre-Servian scholia. And we shall know more when the tardy French printers make the British Academy edition of the Liber Glossarum (or Glossarsum Ansileuli) available to scholars. W. M. Linpsay. Th libens an that the a shorte indicati further. Pr through are coIr sion, an in Italy which ] sister, i the port yoke (tI or posst as Sir J under t attentio ‘hs < p. 215= sel. 9289 that the : 2 Arch cf, inscri 24337 (st palms), 2 24346, 24 the sanc 24169. Au(gustt) ow, galbinus ably referred imt may have re pinnarum, u sex pedum ; mevops with quite unlike nation. ook material this scholium atque alere p, but not as whether the us exemplar, y some later r witness to orrowed the he galbeolus of the truth. s of Servius yself do not nd we shall lition of the LINDSAY. INTRARE SVB IVGVM. Tue following inscription was found at Neferis in North Africa :? stay crescent moon star star SATVRNO AVG, SA CRV C., MEMMIVS PVDENS ~ SACER DOS INTRAVIT SVB IVGV L.A. three pomegranates. That is, Saturno Augusto sacrum. Gatus Memmius Pudens sacerdos tntraut sub 1ugum libens animo. Intvauit sub iugum is peculiar to this text, but it is reasonable to suppose that the plain intyvauit used in other inscriptions of priests of the African Saturnus is a shorter equivalent. Professor Schulten has suggested that some symbolical act indicating the assumption of priesthood is thereby indicated.2 We may go a little further, Probably the rite thus commemorated is one of those ceremonies of passing through a hole or divided object, such as a cleft tree, or between two posts which are common as a means of securing freedom from disease or from demoniac posses- sion, and are associated with initiation and rites de passage in general. Traces of this in Italy are perhaps to be found in the sovorium tigillum at Rome, by passing under which Horatius was thought to have been purified from the guilt of murdering his sister, in the return of the Roman army from the field under the archway known as the porta triumphalis, and in the Italian custom of sending conquered armies under the yoke (these last are passing from one status to another, as Warde Fowler explains, or possessed of malignant and hostile powers of which it is desirable to deprive them, as Sir James Frazer puts it). Confirmatory evidence as to the meaning of passing under the yoke has recently been adduced by Professor Halliday. He has drawn attention to a similar Tartar practice of going between two javelins set beside two with that intention, cf. ].H.S. XLV. 92. 3 Cf. A, Van Gennep, Rites de Passage, p. 22; Roscher, Lex. I1., p. 21 ;J.G. Frazer, G.B.3 X1., p. 168 sqq., 193 sqq.; W. W. Fowler, Roman Essays and Interpretations, p. 708qq., 152. On the tigillum see H. J. Rose, Mnemosyne, N.S. LIII. (1925), 1 P. Gauckler, Bull, soc. antiqg. France, 1905, p. 215=C.JI.L. VIII. 24034= Dessau, Inscr. lat. sel. 9289. Mr. D. B. Harden suggests to me that the star at the top may be a solar disk. 2 Arch, Anz., 1909, p. 206. For intrauit alone cf. inscriptions from Djebel-Djellud, C./.L. VIII. 24337 (surmounted by stars, crescent moon, and palms), 24344 (stars and crescent moon), 24341, 24346, 24347 (all with similar symbols), and from the sanctuary on Mons Balcaranensis 24168, 24169. On 24346 there is added fro sal(ute) Au(gusti): any religious act might be performed P. 407 Sqq. 4 Folklore, 1924, p. 93 sq-, to which Dr. M. Cary kindly drew my attention. R. Wiinsch, Hessische Blatter fiir Volkskunde, VIII., p. 129, explained the custom of making soldiers pass under the yoke as symbolical of subjection, 108 A. D. NOCK fires, under strips of buckram tied to a cord bound to the tops of the javelins. This rite was used to purify the relatives of a man who had recently died, or those who had been near anyone killed by lightning ; it was also employed to remove any noxious properties attaching to ambassadors from abroad and to the gifts which they brought. Here there can be no doubt of the significance of the spears and the horizontal bar above them. An alternative explanation, which seems less attractive, of intrave sub iugum would be that the god here given the name Saturnus is a Baal of primarily celestial character, and the iwgum a cosmic symbol: the stars and crescent moon accompanying the inscription may signify sky, the pomegranates earth.’ It should, however, be remembered that the Saturnus of Roman North Africa is more concerned with the fruits of the earth than with the sky,? though he is lord of the latter also, and that is hazardous to attach a special significance to symbols so com- monly given to him. In either case the priest makes a solemn entry under the yoke. This, as has been said, is probably a way of entering on the priesthood.? Elsewhere we know of a ritual entry as a symbol of initiation, in the obscure custom called é¢uBarevew at Claros.* The reception of priesthood may well be regarded as a form of initiation: an inscription found in Africa proconsularis runs thus, Haos Augusto sacrum. L. Lepi- dius Primulus sacerdos hoc loco initiatus aram posuit, uotum soluit.5 Such initiations are common among primitive peoples ; thus a newly-made priest in the island of Nias must creep through a cleft in a tree.6 In classical antiquity most priesthoods did not, it seems, involve any solemn ordination or investiture : at Rome the vex, the chief flamines, and the augures were admitted by imauguratio, but inauguratio means no more than a formal declaration of the deity’s satisfaction with the choice of person,’ and there is no indication that any rite was performed on the new priest involving a change of state, though the office was normally held for life. In Greece, on the other hand, there is evidence that the priests of certain cults were admitted by a form of initiation. At Cos the priestess of Dionysos and the priest of an unknown deity, at Amorgos the priestess of the Mother of the gods, at Chalcedon the priest of 1 Such astral symbols are common in North Africa, as for instance on funerary monuments (J. Toutain, Revue des Etudes anciennes, XIII., 1911, p. 165sqq.). A. B. Cook regards the dokana of the Dioscuri and also the arch of Janus, the trium- phal arch, and the figillum sororium as cosmic symbols (Zeus I., p. 766 sqq. ; II., p. 354 sqq.): on the whole the probabilities seem to be against this view. 2 Cf. Wissowa in Roscher, Lex., [V. p. 442, 1. 63 sqq., and Toutain’s discussion of the god’s character, Cultes paiens dans l’Empire romain, Premiéve Partie, I1I., p. 18 sqq. 3 A. Merlin, Notes et Documents, IV., p. 24, connects our intrauit with a Punic text found in a sanctuary of Baal and Tanit near Siagu and recording its dedication (1. 4, ‘Entrérent ces dieux-la dans ces sanctuaires-l4 au mois de Mopha de I’année suivante’), and supposes that the priest’s entry was thought of as parallel to the god’s. This does not explain sub iugum. 4 Cf. Ch. Picard, Zphése et Claros, p. 303 sqq. ; W. M. Ramsay, B.S.A. XVIIL., p. 46 sqq., finds the same conception in the temple of Men Askaenos at Antioch, but this is not certain. éuBaredw may have a nuance of another mean- ing, ‘enter into possession’ (Stephanus, Thes. ed. Paris III., p. 8088; E. H. Minns, J7.H.S. XXXV., p. 58, n. 121). Picard, op. cit., p. 271, mentions other instances of a ritual entry or mode of entering. 5 C.I.L. VIII. 16759. Gsell, Inscr. Algér. I, 1179 (Haos is known to us from this single inscription. Cumont, Pauly-Wissowa, VII., p. 2363.59 regards him as a native Numidian deity). Cf. C.1.L. VIII. 15028, Sentius Citus sacerdos profesus est lib(enti) animo. 6 Cf. Hubert Mauss, Mélanges d’Histoive des Religions, p. 131 sqq.; G. Landtman, The Origin oy Priesthood {Ekenaes, 1905), p. 123 sq. : for the Nias cf. G.B.* XI., p. 174. Absolute consistency is not to be expected in such matters; thus among the Lapps a baptismal rite is performed on would-be shamans, although their powers are often thought to be hereditary (W. Bonser, Folkiore, 1924, p. 59 Sq.), and the Lyngdoh, or sacrificers of the Khasi in Assam, sometimes commence their sacrificial career with a solemn sacrifice, like Greek elci.rjpa (Ph. Legrand, Dar.- Saglio, IV., p.939), sometimes without any special ceremonial (P. F. Stegmiller, Mitt. anthrop. Ges., Wien, LIV., 1924, p. 213 sqq.). 7 Cf. Wissowa, Religion und Kultus dey Rémer?, p. 490. Asklepios should pr sacred fu: It m directly ¢ a list of | Such ac phrase sa enable tl that Arte his on Cr associati: mus Doli deeper s! thing lik CLAR 1 Ditter Paton-Hi Syll8, 104 initiation this inscr: unknown omelcOn €7 2 Syll. Lexiph, 1 XVI., Pp. commem doubtful. Egypt th a kind o elins. This r those who remove any s which they ars and the S attractive, is a Baal of und crescent -earth.? It rica is more lord of the ols SO com- This, as has we know of rBareverv at ' Initiation: m. L. Lepi- ‘lations are nd of Nias sthoods did 7x, the chief ns no more erson,’ and - Involving sce, on the | by a form own deity, > priest of 1.S. XXXV,, 71, mentions or mode of nscy. Algéy, | this single a, VII., p. idian deity). tus sacerdos ‘Histoive des The Ovigin oy q. : for the consistency itters ; thus ; performed eir powers W. Bonser, syngdoh, or sometimes 2 a solemn rand, Dar..- any special throp. Ges., ley Romer?, INTRARE SVB IVGVM 109 Asklepios, at Pergamon the priest of Dionysos Kathegemon were thus admitted :* we should probably add the priests of Zeus Panamaros at Panamara in Caria, and the sacred functionaries at Eleusis, of whom Lucian says ¢£ obrep wowwOnoav,? It may be noted that we find also in North Africa the belief that the priest is directly chosen by Saturnus for his service. a list of priests headed— This is a reasonable inference from sac]JERD. QVOS INPOSVI[T njON. IVN. IPSE ASCENDIT.® Such a choice might be thought to be conveyed in a dream, as we may infer from the phrase sacerdos somnio factus.* So in Greece the choice of priests by lot was held to enable the deity concerned to select his minister.° Further, we are expressly told that Artemis chose her priestess on Patmos, Zeus of Panamaros his priest, Asklepios(?) his on Cos, while a Vestal virgin is described as a dits electa, and the members of an association for the worship of Juppiter Dolichenus as quos elegit Iuppiter optimus maxt- mus Dolichenus stbi sevurre.® Whether this African entry under the yoke was given any deeper significance under the Empire, so that intrauit sub tugum came to mean some- thing like a phrase in Apuleius, ministerit iugum subi uoluntarium,'’ we cannot say. CLARE COLLEGE, CAMBRIDGE, 1 Dittenberger, Syl/?, ro1rz. 20 (reA®) and Paton-Hicks, Inscriptions of Cos, 29 (rederd) : Syll3, 1047. 16 (ré\eorpa means ‘fee for this initiation’): Syil3, 1009. 21 (dvariOnu is used in this inscription as in 1o11. 12 of the priest of an unknown deity) : Or. gr. inscr. sel., 331, 20 (karte- omelaOn émi ra lepa). 2 Syll3, goo. 49 (uvoraywyhoas Tov ieph) : Lucian, Lexiph. 10. Dittenberger’s explanation (Hermes, XVI., p. 175) of Pindar’s "Ev@poncopol as odes commemorating the installation of priests is very doubtful. For the belief in Graeco-Roman Egypt that the would-be magician should receive a kind of initiation cf. Th. Hopfner, Griechisch- A. D. Nock. digyptischer Offenbarungszauber, II., p. 14, § 28 sq. 3 Gsell, Inscr. Alg., I. 3018 (with his note on ascendit), 4 Dessau 4445 (a dedication to Saturnus Bal- caranensis): for primitive parallels cf. Landt- man, op. cit., p. 96sq. 5 So I.G. XII. 3. 178 (Astypalnea) dodery Gels iwd Tras Oeod Sid Tod KAdpov lepeds: cf. J. Martha, Les Sacerdoces athéniens, p. 30 sqq. 6 Kaibel, Epigrammata gvraeca, 872. 2: H. Oppermann, R.G.V.V., XIX, 3, p. 54 sqq.: Paton- Hicks, 103: Dessau, 4935: Dessau, 4316 (at Rome). 7 Met. XI. 15. SUMMARIES OF PERIODICALS. LITERATURE AND GENERAL. American Journal of Philology. XLVI. 3. 1925. P. Haupt, Philological and Linguistic Studies. Notes on The Historical Nucleus of the Odyssey: Transposition, Assimilation, and Conformation; The Origin of Semitic Case-endings ; The Odyssey and The Gilgames Epic, etc. H. Collitz, Gothic Siponeis a Loan Word from Greek. Argues that siponjan and siponeis are to be identified with ovyroveiv and cipurovos. E. Reiss, Studies in Superstition and Folklore. VII. Homer. Discusses the magic circle, ancestor worship, and the evils of childlessness, Calypso as a chthonic deity, Circe as a witch, the power of the dead, rites of initiation, the sanctity of the hearth, the Phaeacians, and the meaning of the words GéAyewv, kabdrrer Oar, and ddppaxov. S. E. Bassett, The Laocoon Episode in Quintus Smyrnaeus. B. suggests that Q.’s main sources were Bacchylides and the Epitome of Apollodorus, while certain features in his narrative may have been derived from Sophocles, B.E. Perry, On Apuleius’ Metamorphoses II. 31 ¢o III. 20. P. seeks to show that, while the fight with the wine-skins was an integral portion of the original story, the episode of the Risus festival was an addition of Apuleius’ own. W. N. Bates, Notes on the Dating of the Homeric Poems. Recent archaeological discovery goes to date the sack of Mycenae at the end of the twelfth century B.c., and the Homeric poems must be earlier. The Cnossus tablets in ‘ Linear Script B’ are written in ‘Achaean ’—that is, Homeric Greek, J. Elmore, Recto Vultu and Recta Facie in Juvenal. These phrases describe the abnormal strain and rigidity of fixed attention. XLVI. 4. 1925. C. W. Mendell, Vt Clauses. I. Deals with clauses (i.) in which the w¢ clause is the whole sentence; (ii.) in which it is injected into an otherwise complete sentence, H. L. Ebeling, The, Persians of Timotheus. The plot of the poem was based upon the Aeschylean account of the battle of Salamis. J. E. Harry, The Helena of Eunpides, A series of conjectural emendations: v. 122, eZdov jv for «iddunv; v. 125, ereias for efras; v. 154, y’ €ro6’ ws for reroOds ; v. 292, yvwoOy for cwOy ; v. 277, Oxp’ hv for dye; v. 297, joOev for éoriv; v. 302, épar(d) for apr’; v. 389, ais & ws for meoes; v. 679, ri & for 748’, and ro 8 for ravd’; v. 866, bes pev for Oeopdv; v. 936, kata Ady’ iv for xaterpdyn; v. 961, marav for rdOw; v. 1374, Sy Tav Hpracev for byt avipracev; V. 1535, ois éevnv for eis ev fv. In v. 1597 the traditional text is defended. Catharine Saunders, Cremation and Inhumation in the Aeneid. In his descriptions of burial rites Vergil, with his usual accuracy, thought of Trojans and Arcadians as practising cremation, and the South Etruscans and the bulk of the Latin allies inhumation. H. Collitz, Gothic Barusnjan. The first element of this word (which is used in 1 Timothy V. 4 to render evoefeiv) is a transliteration of the Greek Bapis, used in its post-classical sense of ‘ dignified’ or ‘ of importance.’ Classical Philology. XX. 4. 1925. G. L. Hendrickson, Verbal Injury, Magic, or EvoticComus ? H. examines the use of occentare ostium by Plautus, and concludes that it means ‘to serenade at the door,’ Greek xwpdfew éxi Ovpas. It has no magical significance. Q. Vetter, Quelgues Remarques suv le Papyrus Mathématique No. 621 de la Michigan Collection. A discussion of the Egyptian method of division and treatment of factors. J. A. O. Larsen, Represente powers Represer strength. unit free L. produ Athens 4 Moreturr the first welcomi! review i! Gnomon, F. E. R translate from Ge would re Phil F. | bet Plotin to cases’ with it a Scholium view as Egyptiai of Hora Antony son of tk T. Schw and thou the exce K. Rup] forms V (2) Use Ptolemat 100 A.D Pe View. makes 1 LX W. guilty ¢ Dates tl ‘Ovid e example und exes series 0! of the « Travel ; Sternbila means ; O. Vied regardir cal Nucleus > Origin of ollitz, Gothic s are to be nad Folklore. he evils of of the dead, aning of the le in Quintus the Epitome erived from . P. seeks ‘tion of the uleius’ own. s+haeological ry B.c., and ript B’ are and Recta lity of fixed it clause is e sentence, vased upon Helena of nV; VU. 125; On ; v. 277, is 8 ds for dv; v. 936, pracev for 1al text is 1. In his rojans and ulk of the ant of this eration of ‘tance.’ les the use the door,’ » Quelgues discussion ). Larsen, SUMMARIES OF PERIODICALS III Representative Government tn the Panhellenic Leagues. L, deals with the constitution and powers of the Hellenic league of Philip, Alexander, and Demetrius Poliorcetes. Representation of cities and ethnic groups was apparently in proportion to size and strength. Despite the powers assigned to it and to its proédyot, it left to each local unit freedom and autonomy. J. O. Lofberg, The Date of the Atheman Ephebeia. L. produces evidence to show that the é¢7feua described by Aristotle, Constitution of Athens 42, 4-5, existed long before 335 B.c. O. J. Todd, The Authorship of the Moretum. T. considers that, apart from other considerations, the short quantity of the first syllable of abicit (line 96) proves that Virgil is not the author. P. Shorey, welcoming a new German monthly, Gnomon, deals in particular with Housman’s review in the Classical Review of Merrill’s Catullus (Teubner), which is reproduced in Gnomon, D.M. Robinson refers to the use of ofeAioxo: or iron spits as small change. F. E. Robbins would place a comma after 7Aixos in Aristophanes’ Frogs 55, and translate, ‘ Just a little one—a Molon!’ D.L. D. sees in the clumsy borrowings from Georgics I]. a reason for not ascribing Catalepton XIV. to Virgil. P. Shorey would read avris for avryjv in Plato, Philebus 64a. Philologus. LXXXI. 1. 1925. F. Heinemann, Die Spiegeltheovie dey Materie als Korvelat dev Logos-Licht-Theorie bet Plotin. C. Wendel, Kustos-Wiederholungen in den A pollontosscholien. Draws attention to cases where words repeated in a marginal gloss from an earlier Scholtum, to connect with it a supplement or correction, have wrongly crept into the text of the original Scholium. PP. Corssen, Die vierte Ekloge Virgils. A detailed criticism of Norden’s view as put forward in his Die Geburt des Kindes. C. is particularly sceptical of Egyptian influences on Vergil. The poem is an answer to the sixteenth Epode of Horace, and V.’s ideas are Greco-Roman. The child must be child of either Antony or Octavian or Asinius Pollio, and the early identification with Saloninus, son of the last-named, is perhaps the most likely. But V.’s ambiguity is intentional. T. Schwierczina, Fronto und die Briefe Ciceres. A list of coincidences in expression and thought. J. Schnetz, Jordanis beim Geographen von Ravenna. The latter inserted the excerpts himself with a certain freedom, and not always with acknowledgment. K. Rupprecht, Zwei Probleme dev griechischen Syntax. Discusses (1) use of masculine forms where feminine might be expected. Quotes parallels from inscriptions. (2) Use of the neuter participle as an abstract. O. Stein, Zury Datwerung von Ptolemaios’ Geographie. Indian inscriptions enable us to fix P.’s birth as about 100 A.D. W. Bannier, Ein Papyrusfragment aus der Chrontk des Hippolytus = Oxy. P. VI., No. 870. B. Warnecke, ’Avricxnvos des ephesischen Theaters. Vitruu. V. 6, 4 makes it probable that this addition was for acoustic purposes. LXXXI. 2. 1925. W. Nestle, ’Arpaypootvn. Discusses the reference in Thuc. II. 63. T. is not guilty of an anachronism. W. Judeich, Die Zeit dev Friedensvede des Andohides. Dates the speech at the beginning of 392 B.c. F. Wilhelm, Zu Ovid. Ex Ponto 1. 3. ‘Ovid even here shows himself product of the rhetorical school.’ Proofs of this and examples of O.’s borrowing in this poem from earlier poets. Gundel, Texthritische und exegetische Bemerkungen zu Manilius. (Badly mangled in printing.) Discusses a series of passages. C. Hosius, Die litevarische Stellung von Ausons Moselled. Analysis of the conventional element in the three sections of poem—viz. (1) Description of Travel; (2) Encomium; (3) Autobiography. R.Lehmann-Nitsche, Aus ethnologischen Sternbilderstudien. I. Die Jugula. Discusses this appellation for Orion’s Belt. It means a straight yoke. In Varr. de L.L. VII. 50 read imfra quas for inter quas, O. Viedebantt, Forsehungen zuy altpeloponnesischen Geschichte. Examines the tradition regarding Pheidon of Argos, and discusses (in an excursus) the ‘so-called Second 112 SUMMARIES OF PERIODICALS Messenian War.’ T. Stettner, Casstodors Name. Produces fresh evidence that the) name was Cassiodorus, not Cassiodorius. +R. Foerster, Zu den griechischen Physiognos mikern, A short account of new manuscript material. B. Warnecke, Ad histrionum: vitas. Discusses certain improbabilities in the biographies of Greek actors. | Philological Quarterly. IV. 43. 1925. ] H. Klingelhéfer, De Livit Capite VII. 2. (1) A discussion of Livy’s method of! digression. A digression is always linked to what precedes and to what follows by’ recurrent or anticipatory phrases. (2) A defence of the view that L. derives his: information in this passage from the works of Varro. ¥.'3. 1906. : W. A. Oldfather, A Fleury Text of Avianus in Vaticanus Reginensis 208. Contains a good account of the tradition of the text of Avianus, Cornelia C. Coulter, | Boccacto’s Acquaintance with Homer. 7 LANGUAGE. Glotta. XIV. 3, 4. 1925. A. Nehring continues his article on rirag rirjvn and a pre-Greck k-suffix. | He discusses a number of Greek proper nouns commencing with rir-; e.g. Tirupos, | Tirvppos, Tiraia, Tirav, and connects them with Etruscan and the languages of Asia ™ Minor. Similarly Tirag, a child’s word, meaning ‘ Dada,’ is of Asiatic origin, and its © -k suffix is Asiatic, not Indo-German. There follow reviews of the literature of 1922 | and 1923; Greek, by Paul Kretschmer; Latin Grammar, by A. Nehring; Latin} Syntax, History of Language and Metric, by W. Kroll. G. Maresch writes on The} Name of the Tyndaridae, which he connects with Etruscan tin- ‘ Zeus-Jupiter.’ The | Tyndaridae were the sons of Zeus, and the later Avécxovpor is a Greek translation of 7 their name. Paul Kretschmer writes on The Protindogerman Sivatum, Protindogerman | being the stage the language had reached when Greece was invaded by a pre-Greek, © but Indo-German, race. We may connect it with the ‘ Indo-Germanoid’ languages 7 like Lydian, Lycian, and Etruscan. The patronymic -day suffix seen in Tyndaridae © (v. Maresch above) is connected with -toy in Numitor, etc. We note mrddis: Chaldean * Tuspa-patan, Lycian Pitara, Ilrepin (Boghaz Kevi). The +r has no parallel in ’ Indo-German, the true Greek being wéArs. The -88- (-cmos) suffix seen e.g. in various | place-names—cf Kvw(c)cds, Minassos—has come from the early non-Indo-Germanic | peoples. Zeitschrift fiir vergleichende Sprachforschung. LIII. 3, 4. 1925. i W. Krause, Altindische und altnovdische Kunstpoesie. A comparison of the style of | Kalidasa’s dramas (civc. 500 a.D.) and the Scandinavian scalds (civc. 1000 A.D.). Both incline strongly to nominal, as opposed to verbal, forms of expression, and within the verb itself to periphrastic tenses. K. attributes this tendency to the lack of accent” in the Indo-European finite verb. J. Scheftelowitz, Emm urindisches Liquidengesetz. ‘].-E. / in old Indian disappears before »,s: thereby »>z,s>s. A preceding short’ vowel (unless long by position) is lengthened’; e.g. pai ‘hand’ from *palni; vagd ‘music’: tAdw ; sthiind ‘ post’: erdAos, etc. But I.-E. a/ is differently treated, usually ' becoming i/, w/, and in some circumstances iv, av. Initial v disappears before I.-E. 1) al, av, but is always retained before /,r. A ‘ popular’ pronunciation of /, r asi, % is” already at work in the Rigveda, but is only followed by some of the poets; if the} following syllable contains or v,-each becomes vu. E. Lerch, Vossler’s Aufsatze | zur Sprachphilosophie. Vossler examines the fundamental positions of comparative philology from the philosophic standpoint; he complains that the power of the human | will is ignored. M. Lambertz, Jtalo-albanische Dialektstudten. A glossary. | ce that the! on Physiogno.* ad histrionum: rs. s method of] t follows by? derives his} inensis 208, i C. Coulter, | yeck k-suffix. 7 e.g. Tirupos, | ages of Asia | igin, and its 3 ture of 19224 ring; Latin | ites on The} niter.” The} anslation of © indogerman | . pre-Greek, 7 ’ languages | 1 Tyndaridae ” : Chaldean 7 parallel in 7 ¥, in various } o-Germanic | 1925. | the style of .D.). Both} d within the | k of accent | quidengesetZ. | eding short? ‘palnt; vand® ated, usually © before I.-E. 4 rash, v is™ oets; if the} y’s Aufsatze | omparative | f the human |