than in the case of one who was purely Egyptian. And if it seem strange to any one that a Greek hero who is but a renamed Egyptian calf-divinity should become great in the land, let him observe, first, that the calf Apis was but a manifestation in the flesh of a great and eternal god, and, second, that it is more likely that the Epaphos who had been a calf should become a king, than that an Epaphos who was first a king should be identified with a calf. The explanation of the whole matter, I think, is this. The Greeks were always eager to discover, or invent, the sources of things, and human races they always traced back to the gods. It was not unnatural, then, for them to inquire from whom the black race of Egypt was sprung. In searching for the answer they could find but one possible course of descent from their own gods, who were the only gods, and that lay from Zeus through Epaphos who thus became the founder of the race.31 For it must have been their belief that when Io came into the land of Egypt down the course of the Nile, the land had never yet been visited by human foot, and that she brought with her the seed of the gods. from which should grow the Egyptian race.

32

When Epaphos had been established as the progenitor of the Egyptians, it follows inevitably in Greek thought that he should be the founder of cities, and that Libya should be his daughter. But it has not been remarked, I think, that Memphis, the principal city which he had founded, was the center of the district in which the worship of Apis flourished. For in other parts of Egypt a calf was worshiped, but under another name. This fact itself is a significant bit of evidence in support of the organic connection between Apis and Epaphos.

It remains to explain the story of Epaphos and the Curetes, which Maass takes as evidence of the Euboean origin of the myth. For I hold that no real significance can be attached to the sundry names which were later brought into connection with the myth of Epaphos, and that the only explanation that need be offered

31 Cf. Aesch. Prom. 813 f and 846 ff.

32 The sacred name of the river Nile to which Egypt and the Egyptians owe their very existence is Hapi, identical with that of the calf-god. It is not unlikely that some story which told of the Nile as the source of the race should have been attached to Epaphos through confusion of the

names.

for them is that they were the result of the welter of mythology which in Alexandrian and later times confounded all things. The story of the Curetes, however, has more definite substance, and I find it to be another argument in support of the Egyptian origin of Epaphos. In proof of this let me call attention to the two following points. First, though the bull Apis was at the beginning regarded as an incarnation of the god Ptah, yet "with the growth of the Osirian cult the dead Apis became, like the pious Egyptian, one with Osiris, the lord of the other world. His identity with Ptah paled and disappeared before his newer identity with Osiris. At first he was Osiris-Apis, guardian of the necropolis of Memphis; then as god also of both Memphis and Egypt in life as well as in death. Under the Ptolemies, Greek ideas gathered round the person of a deity who thus united in himself the earlier and later forms of Egyptian belief, and out of the combination rose the Serapis of the classical age, whose worship exercised so great an influence on the Roman world. ''33

.

as

This identity between Apis and Osiris being established, consider the second point. The myth of Osiris contained the following incident, as we learn from Plutarch.34 Set, the brother of Osiris, and called by the Greeks Typhon, plotted with seventy-two others against Osiris. They got him into their power, secured him in a coffer, and flung the coffer into the Nile. Isis, the sister and wife of Osiris, sought for the body and found it at last in Byblos. Now if we recall the myth of Zeus and the Curetes, and set it beside this myth of Osiris, on the one hand, and the legend of Epaphos and the Curetes, as told above, on the other, it can easily be seen that the story of Epaphos is merely a contamination of the myths of Osiris and Zeus, and was attached to the name of Epaphos in later times.

This theory which I have advanced would be completely overthrown if it could be positively proved that Epaphos was mentioned in Greek legend before the beginning of the seventh century. Such proof, of course, may be offered. But for the facts

33 Sayce, A. H., Religion of Ancient Egypt and Babylonia (Gifford Lectures, Edinburgh, 1903), p. 113.

34 Isis and Osiris. Cf. Frazer, J. G., Adonis Attis Osiris (London, 1906), pp. 212 ff.

as they are known at present, this hypothesis seems to me more tenable than that urged by Maass.35 And besides Maass does not answer in any way the question of the origin of the myth, nor explain how this myth came to be attached to the name.

Herodotus, then, was not himself identifying Apis with a distinct Greek mythical personage. He was simply reporting the practice of the Greeks whom he saw in Egypt, and who used their own Greek name for the Egyptian god. They believed, of course, and Herodotus believed, that the name and the myth of Epaphos were both truly Greek; his Egyptian origin must have been forgotten for a hundred years and more.

NOTE.-Without entering into any discussion of the very puzzling myth of Io, I wish to call attention to one matter which may serve to throw some light on the tale. In the Prometheus,36 Aeschylus says that an oracle told Inachos to drive his daughter out of his house and out of the land to wander at large (ǎþeтov) at (èπí with the dative) the farthest limits of the country. This driving a cow, the animal sacred to Hera, beyond the borders, has very much the appearance of an aetiological myth to explain some old expiatory rite resembling the Athenian pharmakos ritual and the Hebrew ceremony of the scape-goat. The word aperоv, as Wecklein observes (ad. loc.), is used technically of animals which are allowed to wander at large in the enclosure of the divinity to whom they are sacred. Thus the myth of Io in its earliest form would have ended at the departure of Io from Argos. Later, the story of the banishment, together with the existence of such names as Bosporus and Euboea, would introduce the idea of the wanderings, which would then be extended through pure imagination. Finally Egypt would be recognized as the resting-place of Io as soon as Isis became known to the Greeks.

35 The words of Aelian (nat. anim. xi 10) seem to show that he had practically the same theory of Epaphos as I have advanced: καὶ ̔́Ελληνες μὲν αὐτὸν [Apis] καλοῦσιν ̓́Επαφον, καὶ γενεαλογοῦσιν οἱ μητέρα Ιὼ τὴν ̓Αργείαν τὴν 'Iváxov ("the Greeks call him Epaphos, and in the established genealogy make him the son of Io of Argos, the daughter of Inachos").

36 664 ff.

During the last generation a movement in the criticism of Latin authors has been going on with increasing force, which has emphasized the individuality and personal character of the Latinity of the separate writers. Formerly a standard of perfection was set up, and by this standard of normal Latinity or of perfection of technique authors were tested: the roughnesses were smoothed away, the strange grammatical usages were emended out of the text, logical inconsistencies were removed in various ways, and the vocabulary was purged. Hofman Peerlkamp in his criticism of Horace marked the climax of the method. But in the latter half of the nineteenth century the school of the Young Grammarians insisted upon consideration of psychological laws in syntax, with corresponding variation in individual writers; and Woelfflin and others established to a greater degree the personal grammar, so to say, of many authors. No longer may we dismiss an et in the meaning 'also' with a superb "et pro etiam displicet"; no longer may we change presents to imperfects in order that the ideal of the sequence of tenses may be observed. Anyone who has given close attention to the received texts of Latin authors must have noticed the hundreds of emendations made by the humanists which have been accepted by succeeding editors; in many cases without serious thought, because on the surface the change seemed necessary to bring the reading into harmony with standard Latin. And in the pre-Ciceronian authors this is especially noticeable; with them grammar was unsettled, many words had been newly created or

were dubious in character; sentence structure had not been standardized, periodology was all but unknown. One need only turn over the pages of such a collection as Bruns' Fontes Iuris to comprehend the difference between the ordinary Latin of common life and the artificial Latin of the stylists.

The task of an advocatus diaboli is naturally invidious, and yet in defending some of the readings of the Lucretian codices I have felt that in our desire to establish the text of Lucretius on certain ground it is necessary to weigh anew the evidence that can be brought for the correctness of the manuscript reading, that, over and over again, has been dismissed by scholars from Marullus down with little hesitation. If the attention of Lucretian scholars shall be drawn to the passages discussed and to the principle involved, the object of this investigation will have been attained.

The conservative position in the constitution of texts is unquestionably the ruling one today, although bitterly opposed in some quarters. Rothstein supports readings in his Propertius (1898) that would have aroused Lachmann's haughty wrath, and almost any Latin Teubner of the last fifteen years, as compared with its predecessor, shows the same tendency. For it must not be forgotten that when the reading of the archetype has been firmly established, the burden of proof falls on the innovator; and even if the novel reading is four hundred years old, still it is a change, and as a change must be justified. And emendations of the archetype must be necessary, cogent, and unavoidable, and not merely desirable or pretty or neat; they must not be admitted if by any reasonable possibility the original text can stand. The authority of such great Latinists as Lambinus and Lachmann has deterred lesser men from abiding by the reading of the Lucretian archetype; Munro, who was quite their equal, in his successive editions manifested increasing courage, and had he lived twenty years longer he would have progressed still further. Brieger in 1894 was still more conservative, and Giussani in 1896-98, and Bailey in 1899 continued the movement. The next task in the constitution of the Lucretian text is grave consideration of every reading of the archetype in the light of modern psychological linguistics.