discovery of the doctrine of the Trinity among the people of Peru. A so-called Icona was found answering to the father, a son (Racab), and a holy spirit (Estrua); nor was the Virgin lacking, nor even St. Anne. "All these things are fictions of the Spaniards." But no sooner has Garcilasso rebuked the Spaniards and their method, than he hastens to illustrate by his own example another difficulty that besets us in our search for evidence of myths. He says, as if it were a matter of certain fact, that Tlasolteute, a kind of Priapus, god of lust, and Ometoctilti, god of drunkenness, and the god of murder, and the others, "were the names of men and women whom the natives of that land worshipped as gods and goddesses." Thus Garcilasso euhemerises audaciously, as also does Sahagun in his account of Mexican religion. We have no right to assume that gods of natural departments (any more than Dionysus and Priapus and Ares) had once been real men and were deified, on evidence like the statement of Garcilasso. He is giving his own euhemeristic guess as if it were matter of fact, and this is a common custom with even the more intelligent of the early missionaries.
Another example of the natural difficulty in studying the myths of savages may be taken from Mr. Sproat's Scenes of Savage Life (1868). There is an honesty and candour in Mr. Sproat's work which by itself seems to clear this witness, at least, of charges of haste or of prejudice. The religion of savages, says this inquirer, "is a subject as to which a traveller might easily form erroneous opinions, owing to the practical difficulty, even to one skilled in the language, of ascertaining the true nature of their superstitions. This short chapter is the result of more than four years inquiry, made unremittingly, under favourable circumstances. There is a constant temptation, from which the unbiassed observer cannot be quite free, to fill up in one's mind, without proper material, the gap between what is known of the religion of the natives for certain, and the larger less known portion, which can only be guessed at; and I frequently
