Reviews. 223
between the horns of the Isis-cow. The yellow face of the image of Osiris in the feast in the month Choiak, and the use of a mould of gold, point to the same fact, and we may suggest as an alternative to Dr. Frazer's readjustment of the Egyptian festivals (pp. 263-267) that there were really two festivals which tended, as usual in Egypt, to syncretise, one in November, the season of sowing, and one at the winter solstice, originally a sun-spell like the Mahavrata of the Vedic ritual.
Dr. Frazer remains a convinced adherent of matriarchy, signs of which he sees in the predominance of the goddess in the cults in question, and from which he explains the marriage of sister and brother, long usual in Egypt, and also the legends (p. 28) of incest by kings. In the latter case, it is significant that Erechtheus and Clymenus are among the guilty. The legends, we think, are nothing more than echoes of the cosmic incest of heaven and earth familiar in Vedic and other mythologies. That female deities mean matriarchy is most improbable (see Farnell, Cults of the Greek States, iv., no), and under patriarchy sister-marriage might be regarded as the best form of marriage, in that it gave the greatest purity of blood, and prevented the waste of the family property in the purchase of a wife for the son, or of a husband for the daughter. For Semites and Phrygians, matriarchy remains a most impro- bable hypothesis.
Of the many other problems suggested by Dr. Frazer, it must suffice to note his view of Hyacinth as a hero (p. 207 ; see Farnell, iv., 264), and the theory that the blood used in puri- fication from murder is accepted as a substitute for the blood of the slayer, with which should be compared the much more probable view of Dr. Farnell (iv., 304) that the blood is that of a sacred animal, and so confers community and friendship with the angry god of the earth and lower world. Nor do we think it possible to accept the theory (p. 97) that the Sandon-Hercules of Lydia was a Hittite god, or that this people were akin to the primitive stock of Asia Minor (cf. p. 17). The facts available show similar religious conceptions all over Asia Minor, and probably generally in the Mediterranean area, and the worship should be assigned, if to any people in particular, to
