the thunder-bird, who has actually been shot in Dacotah, and who is familiar to the Zulus; while rain is the milk of a heavenly cow—­an idea recurring in the ‘Zend Avesta.’  But it does not follow because savages believe in these meteorological beasts that all the beasts in myth were originally meteorological.  Man raised a serpent to the skies, perhaps, but his interest in the animal began on earth, not in the clouds.  It is excessively improbable, and quite unproved, that any race ever regarded lightning as the flashes of a mouse’s teeth.  The hypothesis is a jeu d’esprit, like the opposite hypothesis about the mouse of Night.  In these, and all the other current theories of the Sminthian Apollo, the widely diffused worship of ordinary mice, and such small deer, has been either wholly neglected, or explained by the first theory of symbolism that occurred to the conjecture of a civilised observer.  The facts of savage animal-worship, and their relations to totemism, seem still unknown to or unappreciated by scholars, with the exception of Mr. Sayce, who recognises totemism as the origin of the zoomorphic element in Egyptian religion.

Our explanation, whether adequate or not, is not founded on an isolated case.  If Apollo superseded and absorbed the worship of the mouse, he did no less for the wolf, the ram, the dolphin, and several other animals whose images were associated with his own.  The Greek religion was more refined and anthropomorphic than that of Egypt.  In Egypt the animals were still adored, and the images of the gods had bestial heads.  In Greece only a few gods, and chiefly in very archaic statues, had bestial heads; but beside the other deities the sculptor set the owl, eagle, wolf, serpent, tortoise, mouse, or whatever creature was the local favourite of the deity. {118a} Probably the deity had, in the majority of cases, superseded the animal and succeeded to his honours.  But the conservative religious sentiment retained the beast within the courts and in the suit and service of the anthropomorphic god. {118b}

The process by which the god ousted the beasts may perhaps be observed in Samoa.  There (as Dr. Turner tells us in his ‘Samoa’) each family has its own sacred animal, which it may not eat.  If this law be transgressed, the malefactor is supernaturally punished in a variety of ways.  But, while each family has thus its totem, four or five different families recognise, in owl, crab, lizard, and so on, incarnations of the same god, say of Tongo.  If Tongo had a temple among these families, we can readily believe that images of the various beasts in which he was incarnate would be kept within the consecrated walls.  Savage ideas like these, if they were ever entertained in Greece, would account for the holy animals of the different deities.  But it is obvious that the phenomena which we have been studying may be otherwise explained.  It may be said that the Sminthian Apollo was only revered as the enemy and opponent of