MA S TER NEGATIVE NO. 92-80620-14 MICROFILMED 1992 COLUMBIA UNIVERSITY LIBRARIES/NEW YORK as part of the . t» • *» "Foundations of Western Civilization Preservation Project Funded by the NATIONAL ENDOWMENT FOR THE HUMANITIES Reproductions may not be made without permission from Columbia University Library COPYRIGHT STATEMENT The copyright law of the United States - Title 17, United States Code - concerns the making of photocopies or other reproductions of copyrighted material... Columbia. University Library reserves the right to refuse to accept a copy order if, in its judgement, fulfillment of the order would involve violation of the copyright law. A UTHOR : PRESCOTT, HENRY W. TITLE: THREE PUER-SCENES PLACE: [BOSTON] DA TE : 1910 COLUMBIA UNIVERSITY LIBRARIES PRESERVATION DEPARTMENT Master Negative # \t BIBLIOGRAPHIC MTCROFORM TARHFT Original Material as Filmed - Existing Bibliographic Record Restrictions on Use: ia iM Acquisitions NYCG-PT BKS/3AVE Books FUL/BIB ^ NYCG92-B34i02 Record 1 of - SAVE record ID: MYCG92-B341U2 R FYP : a 3\ CC:9668 BLI:arn DCF:? CSC:? MUD CP:mau L:eng INT:? GPC:? BIO PCis PD:1910/ REP:? CPI M'10: OR: POL: DM: RR: OAO NNC|:cNNC luo i Prescott, Henry W. 2^S lu Ihree PULR -scenes in PJautiis, and Lhe distribution of r oiesf hUnicrol or fli J - 260 (.Boston J,|cl910. 3U0 [311-50 p. urn UiMG QO uS-15-92 FRN: SWU: I-IC:? F S 1 : •? COL: MS: EL: AIC: CUM:??? ILC:???? bML: AD:05-i5-92 UD:05-15-92 II:? UEN: BSE: TECHNICAL MICROFORM DATA FILM SIZE:__3_5Wfrn^i.^__ REDUCTION RATIO: _ //"^ IMAGE PLACEMENT: IA(UM IR IID DATE FILMED:j^VC:^iv _/__ INITIALS____ ^^^_ HLMEDBY: RESEARCH PUDLICATIONS. INC WOODDRIDGe'ct !■■■ Association for inffomiation and image iManagement 1100 Wayne Avenue. Suite 1100. Silver Spring, Maryland 20910 301/587-8202 Centimeter 1 '' """""" "■MliMilim I I I Inches TTT 5 6 iiiliiiiliiiiliiii tt 7 8 iiiliiiiliiiilii mm 9 10 11 iiiiliiiiliiiiliiiiliii TTT 1.0 i^ 2.8 1^ 1 'I ■ 80 n m ■UUta, 1.4 2.5 2.2 2.0 1.8 1.6 I.I 1.25 12 13 14 15 mm iiliiiiliiiiliiiiliiiiliiiiliiiil I I ITT 5 1 MfiNUFflCTURED TO OHM STRNDRRDS BY APPLIED IMRGE, INC. THREE PUER-SCE^ES IN PLAUTUS, AND THE DISTRIBUTION OF ROLES By Henry W. Prescott Printed from the HARVARD STUDIES IN CLASSICAL PHILOLOGY Vol, XXI, igio I:: b| THREE PU£I?-SCENES IN PLAUTUS, AND THE DISTRIBUTION OF ROLES By Henry W. Prescott THE rule of three actors in the classical Greek drama has recently been severely tested. Such a restriction may well have been imposed upon managers of travelling troupes.^ In the Roman comedies a larger number of actors is required, if our texts represent the acting versions of the plays. Ancient evidence, so far as it goes, is in accord with the requirements of the text. After sifting this external evidence, Schmidt'^ in 1870 arranged possible distributions of parts in the plays of Plautus and Terence; in his arrangements he considered almost exclusively coincident appearances of speaking characters upon the stage. Schmidt's results have been generally accepted as showing the practicability of producing the extant comedies with a small number of actors, varying from three to seven in the different plays. The opinion of conservative scholars, probably, finds expression in Hauler's comment on Dziatzko's cautious statement of the case (Ter. Phormio^^ 34, n. 4) : "Untersuchungen wie von Friedr. Schmidt . . . halt Dziatzko insofem fur wertvoll, als dadurch festgestellt wird, mit wie vielen Personen ein Stiick durchgefiihrt werden konnte .... Daraus auf eine feste Regel und etwaige Selbstbeschrankung der lateinischen Dichter bei Komposi- tion ihrer Stucke zu schliessen, wagt er [Dziatzko] mit Recht nicht."* * Rees, The So-called Rule of Three Actors in the Classical Greek Drama^ 64 ff . * Ueber die Zahl der Schauspieler bei Plautus und Terenz und die Vertheilung der Rollen unter dieselben, Erlangen, 1870. Recent discussions of the theme are limited to Terence, and particularly in connection with the notae personarum ; for references cf. Dziatzko-Hauler, Ter. Phormio^^ 34, n. 4. ^ This scepticism was less conservatively expressed in Lorenz*s review of Schmidt {Philol. Anz. V [1873], 459 ff.), and by Steffen in Acta Soc. Phil. Lips, II, 114 ff. Their contention is that the whole question is invalidated by the fact that we cannot discriminate between the Greek original and the Roman adaptation. I think that it is quite proper to study our texts as they stand, and then discover if the scenes in question are demonstrably Greek or Roman. In the present case the conclusion may seem negative, but such evidence as there is seems to me to point to Roman sources for the three scenes under discussion. Cf. below, p. 36, n. 3; p. 39, n. 2; p. 45 and nn. I, 3. f I 32 Henry W. Prescott Audacious as it may be, the thought will occur to any student of the drama that plays produced by a limited number of actors may reveal in their composition the effects of this limitation, especially if the play- wright is a craftsman of very moderate ability. And even a consummate artist may at times be forced to make concessions which his art cannot conceal. The consideration of the plays from this standpoint is not only not audacious, but essential to a proper appreciation of the question. The audacity consists in drawing conclusions from insufficient evidence, and this is all that Hauler means. Naturally students of the Greek drama have sought confirmation of the rule of three actors in the struc- ture of the tragedies ; ^ perhaps they have seen too much, but, granting the theory, their attitude was justifiable. Schmidt, on the contrary, makes no account of internal evidence in this sense of the term : he does quote Poen. 123, 126 (ego ibo, omabor ; . . . ibo, alius nunc fien void), but one of the two verses is probably the work of a retractator, and at best the passage proves only that the prologus did not require a separate actor ; Schmidt also speculates (p. 9) as to what Plautus could not do — he could not present dramatically the joy of the wives in the Stichus over the return of their husbands, because two actors carried the four roles, nor could he let the audience see the reconciliation of the father and the son in the Bacchides, if one actor played both parts ; such speculation is of negative value, and tends to indulgence in idle fancies. What kind of internal evidence is available for our purpose? On p. 13 Schmidt (accepting the rule of three actors for the Greek drama) remarks that the presence of the chorus in the classical tragedy and comedy facilitated change of roles ; in the later Greek drama, he adds, the chorus disappeared ; more characters have to appear, " und so mehrere Zwischenscenen entstehen"; new actors appear in these "Zwischenscenen," and the other actors have an opportunity to change roles. We now know that the chorus did not entirely disappear from the later comedy, and perhaps traces of it survive even in Roman comedy,* but Schmidt's statement of the case applies as an df pnori description to our plays. A skilful dramatist, if limited to a few actors. » Cf., for example, Elmsley in Class. Jour, VIII (1813), 433 ff. « For survivals of the chorus, and substitutes for it, cf. Leo, Der Monolog im Drama, 43, 44; 44t »». I and 2; 50, n. 6; 59, n. 2. Three Vuer-Scenes in Plautus, and the Distribution ef Roles 33 will so arrange his scenes as to provide for the shifting of r61es, and often without much difficulty. The ease with which it may be ac- complished is clear enough, if Schmidt's thesis is accepted, from the comedies. In fact, a dramatist is so little hampered by the necessity of distributing a dozen roles among half a dozen actors that internal con- ditions need seldom reveal indisputable cases of concession to this economical device. For we may not accept evidence unless it proves that the roles must have been distributed : the mere possibility has been demonstrated by Schmidt. Accordingly, the "Zwischenscenen" must be of a very peculiar character to serve our purpose ; any play given by a complete cast with a single role for each actor may show "Zwischen- scenen" of the same sort as most of those in Roman comedy — for any dramatist will be likely to arrange a certain amount of alternation in the appearance of characters or groups of characters. If, however, we can discover scenes or passages peculiar in the sense that no reasonable explanation of their peculiarities exists outside of the limitation imposed by a small troupe, such evidence will be of primary value. Such evidence, if discovered, will be quite independent of external evidence and of Schmidt's entire thesis. But if the evidence of primary importance is convincing, we may properly take up secondary evidence : that is, we may now assume Schmidt's arrangements of roles to be correct, and proceed to test them. This seems an obvious thing to do ; yet Schmidt made no effort in this direction. He tested them only with reference to coincident appearances, referring occasionally to har- mony of roles. But simple practical tests are at hand. If in a given play Schmidt thinks there were five actors in the cast, and if in a succession of two scenes, one requires five speaking roles, and the other a role not included in the five, the structure must reveal the opportunity given to change roles. If in the same play in a succession of two scenes one requires four parts and the other three parts distinct from the four, Schmidt's theory is immediately tested. Now such conditions do exist in the plays ; the cases are not many in number, but they all corroborate the general theory of a limited number of actors, and either correct or confirm, to some extent, Schmidt's arrangements of roles. It should be remembered, however, that such evidence or tests never carry us beyond the standpoint of Dziatzko. Only the primary evidence, if it is convincing to others as it is to me, will lead us to modify 34 Menry W, Prescott Dziatzko's conservatism as Hauler interprets it. If my conclusions are correct, internal evidence shows that the Miles GloHosus, the Captivt, and the Pseudolus were in their present form intended for production by a relatively small number of actors; Dziatzko^s (or Hauler's) ^^ could have been produced" becomes m these three cases ''must have been produced." Furthermore, the evidence in these three cases shows that the author of any one of these plays in its present form (whether Plautus or not I do not yet care to say) was subjected to a certam "Selbstbeschrankung." This is the important point, not that the roles were distributed among a few actors, but that the present form of the Latin comedies was affected by this concession to economy. The last two acts of the Miles are parts of a harmonious whole. No theory of contamination^ seriously affects their integrity. In them the second and final intrigue against the miles is successfully accomplished. A fictitious wife of Periplectomenus is represented to have fallen in love with the miles ; the new affair necessitates the withdrawal of the former sweetheart, Philocomasium — an issue which satisfies her lover, Pleu- sicles, and the arch-intriguer, Palaestrio. The fourth act carries the intrigue up to the catastrophe ; the fifth act reveals the catastrophe. In the action of the fourth act two circumstances are of importance in our present study. In the first place, Milphidippa, posing as the servant of the wife, is clearly presented as a go-between ; it is she who makes the advances in IV, 2, and brings him a ring from the wife (1048-9) ; in IV, 6 she appears with the wife as her confidante ; in both places her part is important, and her activity the conventional role of the servant or nurse in Hellenistic love-stories. In the second place, an important fact comes out in her words in 1277 ; the miles inquires how he is to gratify his new sweetheart when her husband is still in the field ; quin tua caussa exegit virum ab se, is the servant's answer ; there is, then, no obstacle to the new amour save the presence of Philocomasium. Act IV, 8 removes this obstacle ; m this scene Philocomasium, Pleu- » Cf. Leo, Plant. Forsch., 161 ff., and the references in n. 3, adding Hasper, de compositione Militis Gloriosi, Festschrift d. 44- Versammlung deutsch, Phil, und Schulm.y pp. 335 ff., Dresden, 1897, and Kakridis, Rh. Mus., $9 (i904)» 626. ,. Three Vuer- Scenes in Plautus ^ and the Distribution of Roles 35 sides, Palaestrio, and Pyrgopolinices appear; the first two retire at V. 1353 ; the miles, left behind with Palaestrio, expresses his apprecia- tion of the slave's services, and is with difficulty persuaded not to retain Palaestrio in his employ; Palaestrio retires in v. 1373. In 1373-6 the miles soliloquizes further on Palaestrio' s faithfulness ; he then says : ibo hinc intra nunciam ad amores meos (1376-7). The action thus far makes it quite natural that he should immediately carry out this purpose, but instead of doing so he hears the noise of an opening door : a piur appears, and the following scene (IV, 9) takes place : PvER Pyrgopolinices Pv. ne me moneatis : memini ego officium meum ; ego t nam t conueniam ilium, ubi ubist gentium ; inuestigabo, operae non parco meae. 18» Py. me quaerit illic : ibo huic puero obuiam. Pv. ehem, te quaero : salue, uir lepidissume, cumulate commoditate, praeter ceteros duo di quem curant. Py. qui duo? Pv. Mars et Venus. Py. facetum puerum. Pv. intro te ut eas opsecrat : M8i te uolt, te quaerit teque expectans expetit. amanti fer opem. quid stas? quin intro is? Py. eo. — Pv. ipsus illic sese iam inpediuit in plagas. paratae insidiae sunt : in statu stat senex, ut adoriatur moechum, qui formast ferox, M» qui omnis se amare credit, quaeque aspexerit mulier : eum oderunt qua uiri qua mulieres. nunc in tumultum ibo : intus clamorem audio. The next scene (V, i) requires four speaking characters, the largest number required in any scene of the play ; they are Periplectomenus, Pyrgopolinices, Cario, and a lorarius ; in 1427 a fifth appears, Sceledrus. What is the purpose of the /« The Beures are of special interest in comparison with the length of the pa^ges used by some scholars to prove the distribution of roles in the Greek drama. In C&«. Journ. VIII (.8.3). 433-5, Elmsley remarks: •• It appears from these instances, a,at the recitation of twelve or fifteen trimeter iambics aUowed an actor sufficient time to retire, to change his dress, and return." If there is any force m the correspond- Le (aild I doTot wish to be understood as committing myself to the rule of three actors in the Greek drama), the suggestions of Roemer (/%./»/., 65 [1906], 74) «e weakened by the much greater length of the intervening passage in the Ajax and the AnHgon.. Cf. also Rees, op. cit., 50 ff.. who is " guided by the s.tuat.on m mivd- ual secondly, apart from the style there is nothing in the form and content of the passages that militates against a theory of interpolation by a later hand for the purpose of production under special economic conditions : on the other hand, such craftsmanship as they illustrate seems not inconsistent with the manner of Plautus as evinced in other features of his style and technique. Whether as a stage-hand or an actor in AteUan plays Plautus probably learned the practical side of dramatic composition; his plays show that such practical requirements might very likely be met by him in a far from artistic fashion.' IV There are certain features common to all these puer-scenes : they are short ; they employ a character not used elsewhere in the play ; they do not advance the action ; in two of the three cases they are followed by scenes that require the largest number of actors employed in any scene of the play. Obviously any other scenes that reveal all these features may repay study. Perhaps tiie question has already • Cf. Leo, Dtr Monolog, 59, n. 2. Karsten, however {Mnemos., 31 [1903]. I54). differs with respect to the />«.r.scene in the Pieudolm : " Forma et res prorsus grae- cae sunt. Actionem fabulae paud hi versus nihU promovent, sed neque retment, idcirco scenam. quae respondet querelis Syncerasti in Poenulo. nee Plauto nee graeco auctoreindignam esse censeo." The comparison with the /'..«. 823 ft. .s not .nap- propriate. but positive evidence of a Greek source is wanting; the moUf of 773-4 >s certainly Greek, but who shall say it is not Roman? » Leo, /%«/. /<'"<^*.,64ff., 72ff. ,^ , , . . 3 I mean amply that the incoherences of the style and structure of the plays should lead us to be very cautious in ascribing to other hands parts of the plays that show a weakness in organic structure. The fact that Terence reveals nothing analogous to the /«.r.scen« may point in one of many directions: either to a d.fierence m text- tradWon, or a difference in the matter of dependence upon Greek ongmak, or fincludinE the previous explanation) a difference in personahty and m methods. Until the'style of the /^..-scenes is proved to ^^ .f "-\''°'". ^ '^C' 'J "^' indulge my predilection for the last explanation, with a quahficat.on that the J>u^- scenrin the Pseudolus is more properly suspected than the other two scenes. 4S Henry W, Prescott arisen in the reader's mind: are there not other scenes in which a minor character, not a puer, appears for a short time? This is quite true, but such characters and scenes usually promote the action : for example, Halisca in Cist IV, 2, the advorsitores in Most IV, i, 2. Other scenes fail to satisfy all the requirements, though reproducing some of the features : the choragus'^c^xit in the Cure, IV, i is short, introduces a character not elsewhere used, and is a stationary scene, but the neighboring scenes show that it does not facilitate change of roles, but simply stops a gap between the withdrawal of three characters in III, I and the return of the same characters in IV, 2 ;^ the Lurchio- scene in the Miles (III, 2) satisfies most of the requirements, but it is too long to serve primarily for the change of roles and is sufficiently explained by the contamination- theory ;* the //>^tfA?r^j-scene» in the Rudens (II, i) may very likely suggest that Ptolemocratia (I, 5) and Trachalio (II, 2) were played by the same actor, but the conditions of the three scenes do not make this explanation inevitable ; in the same play the lorani'^3s&z.%^ (821-838) is very suspicious, especially as so many actors are required in the next scene (III, 6), but again the explanation is not inevitable, and such lorarii-sctnts seem rather to be for comic effect;* the convivium-sctnts opening the fifth act of the Asinaria and of the Persa incidentally, perhaps, facilitate change of roles, but they are probably derived from Greek sources and primarily serve other purposes.* » Leo, Der Monolog, 50, n. 6. « Leo, Plant. Forsch., 166 ff. • Leo, Der Monolog^ 44. • There is no evidence of whipping m this scene, but the situation and the threats probably entertained the audience; corporal punishment was clearly a source of comic effect m ancient comedy, and in the case of slaves is included among the elements which Aristophanes pretends to have banished from the comic stage {Pax 743 ff.) in spite of several scenes in his plays that point to the contrary. • In the Asinaria^ V, i, the conviviunt'SctnQ would permit Diabolus to become Artemona, but against this combination cf. Schmidt, op, cii., 16, 22. In the Persa, V, i, the conmvmm-scene would permit the Ftrgo to become Paegnium. This combination is not improbable; cf. Schmidt, op. cit, 32. In both cases the next scene requires the largest number of actore needed in any scene of the play; both scenes are stop-gaps, but the return of Dordalus in the Persa is not motivated as is the advent of Artemona in the Asinaria, Finally, besides serving aU these other purposes, the scenes are. unnecessarily long for a change of roles, and are analogous in content to l^o* in the Old Attic Comedy (Leo, Plaui. Forsch,, 152). Three VuGr-Scenes in PlautuSy and the Distribution of Roles 47 There are, however, three other passages that may properly ser\^e as secondary evidence ; if one grants that the roles were distributed, and accepts a reasonable assignment of parts for these three plays, the scenes in question immediately become necessary and to this necessity their existence may plausibly be referred. The admirable expository scenes of the first act of the Mostellaria are familiar to every student of Plautus. Philolaches, Callidamates, Philematium, Delphium are in the midst of their revelry at the end of the first act when Tranio appears with the news of the father's arrival. Act II, I, therefore, requires five actors, the largest number required in any scene of the play. Schmidt shows that five actors might easily have carried all the roles of the play ; in this case the theory is very plausible, for none of the five characters needed in the introductory scenes reappears except Tranio and Callidamates, and the reappearance of the latter is deferred to the end of the play. The father, Theopro- pides, appears in II, 2 ; the theory requires that one of the five actors in II, I shall take the part of the father in II, 2. If this is the case the structure should reveal the provision made for the change of roles. The two women leave the stage in 398 ; a short conversation follows between Philolaches and Tranio (398-406) in which a very trivial bit of action is developed. This action is in a sense essential : the house- door must be locked that the father may not get in and discover the revellers ; it must be locked on the outside because the revellers are irresponsible, and might interfere with Tranio's plans. But this action hardly requires the attention given to it ; Philolaches might have the key or secure it by giving an order to a slave within the house. Instead of this, Philolaches goes into the house; Tranio soliloquizes in the fashion of the intriguing slave ^ confident of success (409-18) ; d. puer appears with the key (419) ; after a short dialogue, Tranio resumes his monologue (427-30) ; then the father appears (431). It seems reason- able to suppose that this trivial action is developed to allow Philolaches to become Theopropides, a combination for which Schmidt (p. 32) provides ; the passage (408-30 = 23 vv.) is of the length required for such a change so far as the puer-scenes set a standard. The early * Leo, Der Mono/og, 72, n. 12, includes some of these monologues under the head of ** Oberlegung." 48 Henry W, Prescott departure of the women (398) may have enabled one of them to take the part of the puer> But the difference between this passage and the /^^r-scenes is important ; here the puer has few words and contributes to the action ; the monologue of Tranio is the conventional monologue of the arch-intriguer before he puts his plans into operation ; the only suspicious feature is the trivial nature of the action. The technique is less difficult to parallel than that of the /«^r-scenes, and for that reason the evidence is somewhat less positive. The environment of all scenes that require, according to Schmidt's assignments, the entire troupe of ac'tors is likely to test his theory, but only in case a new character outside of the maximum number appears in the preceding or following scene. Schmidt shows that the Mercator could have been presented by four actors. The entire company, there- fore, is required in IV, 4, in which Lysimachus, Dorippa, Syra, and a cook appear at the outset. The next scene (IV, 5) brings a fifth, Eutychus, on the stage. The structure must show some device to meet this situation. The cook and his attendants withdraw in 782 ; Syra leaves in 788; Dorippa leaves before 792; Lysimachus delivers a monologue, 793-802 ; then Syra and Eutychus appear (IV, 4). Such a structure obviously made it possible for either the cook or Dorippa to take the role of Eutychus; Schmidt « with some hesitation combines the roles of Dorippa and Eutychus; in that case at least eleven verses (792-802) intervened for the change of roles; in the other case twenty-three verses (782-802). Both harmony of roles and the struc- ture seem to me to point to a combination of Dorippa and the young lover, Charinus, who appears in 830 ; this leaves the cook and Eutychus for one actor. In this way we secure the obvious fitness of combining a woman's part and that of a sentimental lover; the combination of the audacious cook and the lover seems more difficult. At the same time by this combination of parts we allow plenty of time for changes : the cook has twenty-three verses in which to become Eutychus ; Dorippa has thirty-eight verses (792-829) in which to become Charinus; the greater time required for this change is accounted for by the difficulty » Fritzsche, Quatuor leges scenicae Graecorum poeseos, 29 f!., notes that the role of the puer might have been taken by some one of the actors who appeared in the previous scene, but does not draw any further inferences. * Op, cit,, 56, cf. 30. Three Vu^r-Scenes in Plautus, and the Distribution of Roles 49 of shifting from the role of a woman to that of a man ; incidentally the monologue of Syra (817-29), by this arrangement, has an economic justification in addition to the explanation furnished by the parabasis of Greek comedy and certain passages of Euripides.^ Schmidt has not carefully worked out the important question of supernumeraries. It does not seem consistent with an economic theory to suppose that the parts of the danista, of the fidicina Acropolistis, of the virgo Telestis in the Epidicus were taken by supernumeraries ; all of these characters speak a goodly number of verses ; by his arrange- ment Schmidt brings the number of actors down to four. If we dis- tribute these three roles among the regular actors the troupe need be increased only to five ; and in that case the structure of the play at the beginning of the fifth act becomes intelligible. Act IV, 2 presents three speaking characters; in the course of the next scene (V, i) four additional speaking characters appear; the structure, however, easily provides for a company of five actors. Acropolistis, Periphanes, and Philippa remain on the stage through practically the entire scene (IV, 2) ; after their withdrawal Stratippocles and Epidicus appear and con- verse (607-19) ; then the danisia and Telestis appear (620) ; the danista goes out in 647. The conversation, or rather the two mono- logues and the conversation, in 607-19 (again thirteen verses) do not advance the action or serve any other discoverable purpose than to provide for two of the actors in the previous scene to assume the roles of Telestis and the danista. The possible combinations are the roles of Acropolistis and Periphanes and Philippa^ in the previous scenes with the roles of the danista and Telestis in this scene ; it is natural to assume that the two women, Acropolistis and Telestis, were played by one actor,^ leaving another actor the roles of Periphanes and » This monologue is rejected by Ribbeck {Emend. Merc, Plaut. Spicilegium, 13) and by Langen (^Plaut. Stud,, 312) because of its unfitness and stylistic defects. Leo, however {Plaut. Forsch., 107 ff.), successfully defends it as a survival of "die euripi- deische Klage " {Med. 244 ff., Elect, 1036 ff., Med. 184 ff.). * It is to be noted that, unless there is a pause between the acts, Schmidt's com- bination of Philippa and Epidicus gives Philippa only an interval of six verees (604-9) in which to make the change. ^ This assumption rests on the UkeUhood that the parts of women were taken by one actor so far as possible; in several plays, however, there are too many women to admit of such a combination, or rather, the appearances of the women are such as to prevent this combination. so Henry W, Prescott the danista. This provides thirteen verses for both changes. Finally, it should be noted that the departure of the danista in 647 provides, perhaps, for his return as Periphanes in 666 (648-65 = 18 vv.). Doubtless other scenes of this sort might be discovered.^ I have rejected many, and chosen these three as offering the most satisfactory secondary evidence available. They show that even as a statement of the possibilities Schmidt's arrangements may be improved by a more careful study of the structure of the plays. Such a study would be indeed audacious but for the truth revealed by the ///