persons punished, any more than there is in the endless labor of Sisyphus. The task of filling a leaky vessel is widely known and variously applied in folk-lore, from Grimm's Märchen to Uncle Remus. Such a task would be assigned to the Danaids in Hades when people began to feel that their bloody deed demanded punishment in the lower world. The fact that the same endless task is also assigned to the uninitiated, or to the wicked in general (see Plato, Rep. II., p. 363 E), is another indication that the fastening of it upon the Danaids exclusively, in later times, is only a matter of convention.

III. - Pliny, Pausanias, and the Hermes of Praxiteles.

BY PROF. HAROLD N. FOWLER,

WESTERN RESERVE UNIVERSITY (COLLEGE FOR WOMEN).

IN the Gazette des Beaux-Arts for August, 1897, pp. 119139, is an article by Miss Eugénie Sellers (now Mrs. Strong), on the Hermes of Olympia. In it she attempts to show that the famous statue is not by Praxiteles, but by Cephisodotus the elder. Her arguments led Henri Lechat (Revue des Études Grecques, 1898, p. 207) to say that henceforth it might be more prudent to use the expression "Hermes of Olympia" than "Hermes of Praxiteles," and S. Reinach, Répertoire de la Statuaire Grecque et Romaine, II. i., p. 173, in his note on the cut of the Hermes calls it "Hermès dit de Praxitèle." Such recognition by prominent archaeologists lends the article additional importance and may justify me in using it as the text for a few remarks. It is worth while to add that

Rayet (Gaz. B. A. XXI., 1880, p. 410; Études d'Archéologie et d'Art, p. 68) suggests that Pliny speaks of the Hermes as a work of Cephisodotus.

The two classical texts relating to the authorship of the Hermes are Pausanias, V. 17, 3, χρόνῳ δὲ ὕστερον καὶ ἄλλα ἀνέθεσαν ἐς τὸ Ηραῖον, Ἑρμῆν λίθου, Διόνυσον δὲ φέρει νήπιον, τέχνη δέ ἐστι Πραξιτέλους, and Pliny Ν.Η. XXXIV. 87, Cephisodoti duo fuere : prioris est Mercurius Liberum patrem in infantia nutriens; fecit et contionantem manu elata, persona in incerto est. sequens philosophos fecit.

Miss Sellers wishes to prove: first, that the Hermes mentioned by Pliny and attributed to Cephisodotus is the Hermes of Olympia; second, that the comparative study of the assertions of Pausanias and Pliny must lead us to accept the attribution given by the latter; third, that, in the present state of our knowledge, everything tends to support this result and to confirm the attribution of the statue to the elder Cephisodotus.

It is not my purpose to discuss Miss Sellers' article in detail. It contains many good observations and shows both learning and aesthetic sense, such as we expect to find in her work. But I wish to take up the first and second points which she undertakes to establish, not only because I think she has failed to establish them, but also because it seems to me that, in common with many others who write on archaeological subjects, she argues from insufficient premises.

The proof that Pliny refers to the Hermes of Olympia seems to consist in the fact that Hermes with the infant Dionysus is referred to only in the two passages cited. The tacit assumption seems to be that this was therefore the only famous representation of this group (Gaz. B.A., l.c., p. 122, note). But this is a mere assumption. Pausanias has a definite reason for mentioning the group at Olympia, because he is describing Olympia. No such reason, so far as we know, constrains Pliny. Before it is assumed that he refers to the Hermes of Olympia, it must be proved either that this was better known than other representations or that this one was by Cephisodotus, which is precisely the thing which this assumption is to aid in proving.

Dionysus was a popular divinity, and it is not unnatural that his epiphany should be frequently represented; for this is the meaning of the representation of infant deities. (See Usener, Sintfluthsagen, passim.)

When Dionysus was to be represented as an infant, he was put in the hands of Hermes. Besides the Hermes of Praxiteles several other representations of the same group are known to us. In the Boboli garden at Florence is a rather unattractive Hermes with wings in his hair holding an infant on his right hand and a caduceus in his left. The proportions of the Hermes are heavy, and the shape of the head and the general attitude are such as are associated with Polyclitus. The position of the arm holding the infant does not seem to me perfectly natural; the drapery is arranged in a way quite out of the question for the fifth century B.C., so that the probabilities are in favor of the assumption that a Polyclitan type not originally intended to represent Hermes and Diony

sus has been adapted by some inferior artist and then copied in Roman times, for the extant figure is clearly late work. Miss Sellers, note on p. 122, mentions also a bronze at Roye 1 and representations on gems, seemingly of Polyclitan style.

Another type, the date of which it is hard to fix, is seen on a coin of Pheneus in Arcadia. Here Hermes is apparently hurrying along and holding the infant Dionysus, if it be really Dionysus, almost at arm's length.

Other representations may be modifications of the Hermes of Praxiteles, but one, at least, in the Louvre, J.H.S. III., p. 107, pl., in which the infant seems to be held in the drapery of the elder god so arranged as to make almost a bag on his left arm, appears to be a more or less independent type.

So at least four types of Hermes with the infant Dionysus, or with some infant, for the child's name has little effect upon the type, exist even now. That others, more or less independent, existed in ancient times, is highly probable. That any of the extant types goes back to Cephisodotus is more or less unlikely. In fact, we know little or nothing of Cephisodotus, except that he was an Athenian, and flourished in the fourth century. That Pliny puts him two Olympiads before Praxiteles does not seem to prove that he is his father. The identification of the Munich group with the Eirene and Plutus of Cephisodotus is probable, for Eirene and Plutus are not popular nor frequently represented divinities; but the style of this group, instead of tending to prove that it is by the artist of the Hermes of Olympia, seems to me to prove the contrary. But this is a matter which can be adequately discussed only at great length and with numerous illustrations. In her discussion of it Miss Sellers exhibits great 'ingenuity, but fails to convince me, and would, I think, fail to convince any one who did not before believe as she does. When we consider that the "Polyclitan" type of Hermes and Dionysus exists in several replicas or adaptations, we might even be tempted to believe that it was the most famous type. Then this would be the type referred to by Pliny,

1 Now at Péronne, published by S. Reinach, Gaz. Beaux-Arts, vol. xxiii., 1900, P. 457. See also Rev. Archéol., 1884, II., pl. 4.

if he must be supposed to refer to the most famous type,— and the Attic Cephisodotus would be assumed to have been in his early youth a pupil of Polyclitus. Stranger assumptions than this have been made with hardly more ground to stand on. I do not, however, suggest this even as a possibility.

An argument advanced parenthetically (Lc., p. 138) against the Praxitelean origin of the Hermes of Olympia is the fact that although there are more or less exact imitations of the type among small bronzes, reliefs, and gems, there is no copy in the size of the original. If it were really a work of Praxiteles, or if it had been universally regarded as his work, it would naturally have been copied. But here again we must not assume too much.1

Of the statues mentioned by Pausanias at Olympia, very few, if any, seem to exist in ancient copies. Those for which such existence has been more or less doubtfully claimed are the following: (1) Paus. V. 17. 4, πaidíov dè èπíxрVσOV KÁÐNTAL γυμνὸν πρὸ τῆς ̓Αφροδίτης. Βοηθὸς δὲ ἐτόρευσεν αὐτὸ ΚαρχηSóvios, "a gilded child, naked, is seated before the Aphrodite. The artist who fashioned it was Boethus of Chalcedon." It has been conjectured that this is the original of the boy drawing a thorn from his foot; but there is no reasonable ground for the conjecture. (2) It has been suggested by v. Duhn that the statue of a seated lady in the Museo Torlonia may be a copy of the statue of Olympias by Leochares, mentioned Paus. V. 20, 10; but the marks on the pedestal in the Philippeum seem to show that Olympias was represented standing. (3) The statue of Cyniscus, Paus. VI. 4, 11, is believed by Furtwängler, Meisterwerke, PP. 452-471 Masterpieces, p. 249 ff., and others to be the original of the "Westmacott athlete" and its replicas, as that is Polyclitan, and the position of the feet agree with the marks on the basis found at Olympia.

=

1 There are several figures which may be regarded as more or less accurate imitations of the Hermes, but with the child omitted (see Roscher's Lexikon d. gr. u. röm. Mythologie, I., p. 2414; Reinach, Répertoire de la Statuaire, II. i., p. 173), but these are not to be regarded as copies, and may very well prove nothing more than the general popularity of Praxitelean forms and postures.