I say this more especially with reference to the etymology in Bryant's and other kindred works, which it is now the fashion to set aside with much _hauteur_; and I assert it without impugning in any way the results of modern philological inquiry, extending, of course, over a much wider field than the writers of the last century could embrace. But I do contend, that when the discussion has reference to the common progenitors of the human race, or the incidents of primitive life--for instance, the names of the ark, and what I may call its accessories, the dove and the rainbow[85]--a certain probability of identity may be presumed in such sort that it may chance that the probabilities of tradition must be held to override the conjectures, and in some cases even the conclusions of philology.[86] [85] _Vide_ "Bryant's Mythology," ii. [86] After the exposition of his own theory, Mr Grote says--"It is in this point of view that the myths are important for any one who would correctly appreciate the general tone of Grecian thought and feeling, for they were the _universal mental stock_ of the Hellenic world, _common to men and women, rich and poor_, ignorant and instructed, they _were in every one's memory and in every one's mouth_, while science and history were confined to comparatively few. We know from Thucydides how erroneously and carelessly the Athenian public of his day retained the history of Pisistratus, only one century past; but the adventures of the gods and heroes, the numberless explanatory legends attached to visible objects and _periodical ceremonies_, were the theme of _general talk_, and every man unacquainted with them would have found himself partially excluded from the sympathies of his neighbours."--_Hist. Greece_, i. p. 608; comp. _infra_, ch. xi.