contrary was usual; for, apart from the cases of hair cut to a moderate length and hanging loose-according to the common explanation, in sign of mourning the female figures on these monuments, many of whom must surely be married women, usually have the hair done up in one way or another. Asia (Conze 58 = Athens 767), whose son is at her knees, Archestrate (Conze 290 Athens 722) and Melite (Conze 803 = Athens 720), who are characterized as wives in the accompanying inscriptions, may serve for illustrations. It is unnecessary to labor this point. = Cases of long hair hanging loose are not very numerous on Attic monuments of the fifth and fourth centuries B.C., and among these the cases where no doubt is possible as to the age and status of the wearers are very few indeed. So far as I can see, the persons so represented may be either matrons or maidens. This is certainly the fact in the earlier half of the fifth century, when we find on the one hand Hera and Clytemestra, on the other hand Athena, Oreithyia, and Pandrosus with long, loose hair (F.-R., Pls. 65, 72, 5, 94). There seems to be the same absence of distinction later. The figures in swings (F.-R., Pl. 125 and Abb. 11 in text) are probably girls. But the chief witness on this side of the argument is Athena (Athena Parthenos of Phidias, F.-R., 70, 96, 109, 2, 127). True, the frequency with which Athena is represented with unconfined hair might be due to the inconvenience of wearing a helmet over a chignon. The difficulty, however, was not insuperable, and the maiden goddess could hardly have followed a practice which on earth was confined to married women. Convincing cases of married women with loose hair are hard to find. Theophante on her gravemonument (Conze 309 Athens 1055) may be a case in point, but her status is in some doubt and her hanging locks may mean nothing but the disarray of illness. If the figure rising from the earth on the Eleusinian pelice from Kertch (F.-R., Pl. 70) is Ge, as Furtwängler thought, this would be relevant evidence. The most unmistakable case I can find is that of Medea in F.-R., Pls. 38, 39; but Medea is a barbarian. If we go outside Attica, we find more satisfactory evidence afforded by the head of Hera on coins of Argos, Croton, etc. The foregoing paragraph deals with cases of hair unconfined or at most confined by a band around the head. It remains to consider the various fashions in which the hair is drawn together behind the ears or somewhere lower down, but is not massed upon the head. In the earlier part of the period under consideration a favorite mode with Athena (e.g. on the Orvieto crater, F.-R., Pl. 108) and others (e.g. Europa and Artemis [F.-R., Pls. 114, 115]) is to tie the ends of the long hair into a bunch. This is scarcely found after 450 B.C. We are, therefore, concerned chiefly with cases where the hair is drawn together at the back of the neck or thereabouts. This is the coiffure of the "maidens" of the south porch of the Erechtheum. It is given to Athena, as on a relief at the head of a decree (Bulletin de correspondance hellénique, 1878, Pl. XII) and two reliefs in Le Bas (Voyage archéologique, Mon. fig. 47, 1, and 48, 1); also on vases (F.-R., Pls. 40, 69); not to speak of statues of presumably Attic origin, like the Athena of Velletri. On the grave-reliefs, among the figures whose hair is thus dressed, Eurynoe (Conze 1021), by reason of her size, and the two who hold their dolls (Conze 817, 880) are certainly girls. Mynnion (Conze 896=Athens 763) is a probable case. On the stele of Lysarete (Conze 755) the figure at the extreme left, not full-grown, has long hair, perhaps braided. Within the limits set to this inquiry I have not found an unquestionable instance of a matron with hair similarly arranged. I am inclined to think, therefore, that this arrangement was adopted principally by unmarried girls. In view of the Florentine statue of Niobe, cited by Mr. Washburn, and the figure of Mnemosyne (?) on the Chigi relief (Römische Mittheilungen, 1893, Pls. II, III), works of unknown provenience, I do not dare to make a more sweeping statement. F. B. TARBELL LYSIAS 19. 22 καὶ τοῦ ἀδελφοῦ τοῦ ὁμοπατρίου ἀποκειμένας παρ' αὐτῷ τετταράκοντα μνᾶς εἰπὼν κατεχρήσατο. This son of Nicophemus and half-brother (suomaтpíov) of Aristophanes is not elsewhere referred to in the speech, but there is abundant evidence to prove that there was no such person. In § 36 we read: eri de paívovrai (Kóvwv καὶ Νικόφημος) οὐδὲν πώποτε διενεχθέντες, ὥστ ̓ εἰκὸς καὶ περὶ τῶν χρημάτων ταὐτὰ γνῶναι, ἱκανὰ μὲν ἐνθάδε τῷ ὑεῖ ἑκάτερον καταλιπεῖν, τὰ δὲ ἄλλα παρ' αὑτοῖς ἔχειν· ἦν γὰρ Κόνωνι μὲν ὑιὸς ἐν Κύπρῳ καὶ γυνή, Νικοφήμῳ δὲ γυνὴ καὶ Ovyarnp. The other son is evidently neither in Athens nor in Cyprus, and it would seem that he had no share in his father's property, for "the rest" (rà ala), that is not given to Aristophanes, is held in Cyprus. Compare with this the calculation in §§ 42-43 that Aristophanes had expended about fifteen talents, and the closing words (§44): kaì ov πрoσλoyičóμela oσa αὐτὸς ἐν Κύπρῳ ἔσχε Νικόφημος, οὔσης αὐτῷ ἐκεῖ γυναικὸς καὶ θυγατρός. Again, take these words from § 12: ἐδεήθη δοῦναι τὴν ἐμὴν) ἀδελφὴν αὐτοῦντι τῷ ὑεῖ τῷ Νικοφήμου. It may be fairly argued that, if Nicophemus had two sons, the proper expression here would be ̓Αριστοφάνει τῷ Νικοφήμου. Further evidence is to be deduced from the fact that on the death of Aristophanes his children do not become the wards of his brother, but their mother's family is "compelled to take care of them" (vayκaσμμévoɩ тpépeiv, §9; cf. §33), and the speaker says that he is "deprived of relations by marriage" (σTepnμévoi kydeσtŵv, § 9), and talks of his sister's dowry as a loss suffered by his own estate, whereas, if the children of Aristophanes had a paternal uncle, their mother's dowry would naturally have been held in trust by him for their benefit. And, finally, it should be recalled that the chief aim of the speaker is to prove that the estate of Nicophemus and Aristophanes has been fully accounted for. He claims that the outlay of fifteen talents by Aristophanes, together with the presumably larger amount (§ 37) retained by Nicophemus for his own use, make up a sum far in excess of what they could have been expected to possess, considering that Conon had left but forty talents (§ 40). If, then, there had been a second son, to whom a proportional share would naturally have gone, the speaker would not have failed to utilize the fact to strengthen his argument. The sum of forty minas, said (§ 22) to have been deposited with Aristophanes by his brother, is exactly the amount of his wife's dowry; and, since the dowry did not become the property of the husband, the phrase ἀποκειμένας παρ' αὐτῷ could be properly applied to it. It will take no great change in the text to rid us of the suspected brother. Instead of τοῦ ἀδελφοῦ τοῦ ὁμοπατρίου ἀποκειμένας παρ' αὐτῷ τετταράκοντα μνᾶς Ι suggest that we read τῆς ἀδελφῆς τὰς ὑπὸ τοῦ ἐμοῦ πατρὸς ἀποκειμένας. Supposing τὰς ὑπό: to have dropped out and ἐμοῦ πατρός corrupted to ὁμοπατρίου, the change of τῆς ἀδελφῆς to τοῦ ἀδελφοῦ would naturally follow. This change not only removes the brother, but gives us a reasonable explanation of the loss of the dowry. Dowries were not subject to confiscation, and were usually secured by mortgage on the husband's estate. The speaker twice (§§ 9, 32) refers to the loss of the dowry, but does not refer to the illegality of its confiscation, though he does say that the children have been illegally deprived of their father's property (§ 8). It is not improbable that Aristophanes induced his father-in-law to let him venture his wife's dowry, relieved of mortgage, since he required control of all his real property to give security to the friends from whom he borrowed (τοὺς φίλους ἔπειθε δεόμενος καὶ ἐγγυώμενος). In § 22, immediately following the clause in question, we are told that Aristophanes borrowed seven minas from his father-in-law and "this also he took and spent" (ò dè Kaì TAÚTAS λαβὼν κατεχρήσατο). It is not pressing the point too far to suggest that the kai before Taúras indicates that the forty minas had also come out of the speaker's pocket. Compare the connection of these two sums again in § 32: ἐθέλομεν πίστιν δοῦναι . . . . μηδὲν ἔχειν τῶν ̓Αριστοφάνους χρημάτων, ἐνοφείλεσθαι δὲ τὴν προίκα τῆς ἀδελφῆς καὶ ἑπτὰ μνᾶς, ἃς ᾤχετο λαβὼν παρὰ τοῦ πατρὸς τοῦ ἐμοῦ. A request for permission to use the dowry would be likely to precede one for a further loan. Before kaтexρýσato in our sentence X has cirov, which has been variously corrected. I am inclined to read ἡμᾶς πείθων (cf. τοὺς φίλους ἔπειθε), the ἡμᾶς being easily lost after μνᾶς. The failure of the speaker to emphasize the illegality of the confiscation of the dowry suggests that the lost speech of Lysias, cited by Harpocration—Κατ' Αἰσχίνου περὶ τῆς δημεύσεως τῶν ̓Αριστοφάνους χρημάτων-Was 1 The number of omissions in this oration is very large. For vó after keîμaι and various compounds cf. Dem. 24. 62; Isocr. 4. 168; Thuc. 7. 77; Plato Apol. 30 E. If the emendation of § 22 is correct the antecedent of as will include Tǹν #роîка. made in an attempt to save the dowry and the seven minas. In that case the καὶ πρότερον πρὸς τοὺς συνδίκους of § 32 will refer to the former trial and not to the preliminary hearing (áváκpiois) of the present one. LYSIAS 18. 14 πάντες γὰρ εἴσονται ὅτι τότε μὲν χιλίαις δραχμαῖς ἐζημίωσε τὸν βουλόμενον τὴν ἡμετέραν γῆν δημοσίαν ποιῆσαι κτέ. Among the numerous emendations of this section I do not find that anyone has ever proposed ἐζημιώσατ' ἄν, and yet the simple insertion of an av removes many difficulties. In the first place it explains how nμiwoar', which is certainly right, was corrupted to nuiwoe, for the Tay would be easily dropped before róv. This corruption would not occur so easily before avrov, and, indeed, if this first trial is purely hypothetical, it is no longer necessary to have a reference to Poliochus in this clause, and we can retain the Tóv of X. The point of the whole passage from § 13 to § 19 is that the Athenians, by confiscating the property in question, will break their oath μὴ μνησικακεῖν (§ 19). This is the oath referred to in the περὶ ὧν ὅρκους ỏμwμókaтe of §13, and the â dè avroîs ¿yndíσaole of § 15. In the closing words of § 14 Lysias, indeed, speaks as if the assumed former trial were an actual one, but that is a mere trick, of which he is quite capable. In defense of παρανόμως φεύγοντος τοῦ αὐτοῦ ἀνδρός he could claim that the case had already been decided once when the citizens voted to forget past wrongs. If the insertion of av is right, and there were not two trials, the Toû avrov ȧvdpós cannot possibly refer to Poliochus. We thus get rid of the possibility that the speech was made in a yрapn Taрavóμwv against Poliochus, as Thalheim still assumes in his Argumentum (p. xliii), and the σúvdikoi of § 26 will be the presiding officers. And, finally, we are relieved of the very improbable assumption that a man, who had already been fined 1,000 drachmas for an illegal act, would venture to bring a second suit, which could be attacked as doubly illegal, because it was contrary both to the decree un μvnoɩkakeîv and to the law against a second suit about the same matter. CICERO'S DE SENECTUTE, §§ 10 AND 37 A. G. LAIRD In Cicero's De senectute 10, the Brussels MS (Bruxellensis 9591), tenth century, reads virtus gravitas and furnishes an explanation of a corruption which has appeared in most of the other MSS. Virtus was originally a gloss on gravitas which has here crept into the text without the change to gravis made by the scribes in other MSS to emend the sentence. This corrected reading, virtus gravis, also appears in A (Ashburnhamensis) and L (Leidensis), the other members of the family to which Br belongs, and represents, as stated above, the variation of their scribes from the original virtus gravitas. In E (Erfurtensis), twelfth century, the gloss was inserted after gravitas instead of before, but the same subsequent change has occurred. A confirmation of the fact that virtus is a gloss is found in the reading of the second hand of V (Vossianus), thirteenth century, where virtus is not put into the text but occurs as a variant upon gravitas. Later, in MSS which had retained the original reading gravitas, as in H (Harleianus), eleventh century, the false reading virtus gravis was also inserted because found in other MSS and not for the same primary cause. In § 37 of the De senectute, the reading of Br, illa domos patrius disciplina, is again indicative of the reading of the archetype of the other two members of its family, A and L. For the corruption domos, so easily formed from domo+mos, next gave way to domus as domos would not construe. Then, in order to avoid an apparently false agreement with domus, the patrius was changed to patria P2 (according to Dahl), patris A2, or patri A1 and L', that it might agree with or be in some subjective relation to disciplina, the et necessarily disappearing for this purpose. We thus account for the reading of A1, illa domus patri disciplina, while the repetition of domus by L' is probably a blunder of its own scribe. The et was probably dropped from Br, as in the great majority of the other MSS, through the failure to observe the plural subject and a desire to construe disciplina as ablative. This may, however, have been the initial stage of the corruption in all three MSS. The reading of ER, illa domo mos patrius et disciplina, was thus undoubtedly the reading of the archetype of AL Br and, though not necessarily the correct reading, yet furnishes the best tradition in this disputed passage. I propose shortly to publish a complete collation of Br with an indication of its various affinities. WASHINGTON UNIVERSITY GEORGE R. THROOP "JUSTICE IN THE AGE OF HOMER" In his examination of the much-discussed trial-scene in Homer's Iliad xviii. 497-508, Professor Bonner ("Justice in the Age of Homer," Classical Philology VI, 24 ff.) rightly concludes, I think, that the two talents which lie before the judges are the stakes contributed by the litigants (one talent each), and are "to be paid over to the one of them who shall plead his case before the court most effectively"-that is, win the verdict (vs. 508). Professor Bonner has also not disregarded the earlier suggested parallelisms with procedure under primitive Roman law. But his concluding sentence (p. 30) is: "In effect the wager [i.e., the two talents in this specific case] corresponds to the damages which according to Homeric practice usually accompanied restitution and redress." That is, the beaten litigant is mulcted in penal damages to the amount of one talent, presum |