This was, until comparatively recent years, the received opinion concerning mythology, and it is one which tacitly keeps its place in the writings of many scholars, especially of those who have been brought up almost exclusively upon the study of classical languages and classical religions: for it is only after a wide study, and a comparison of many different religions in many different stages, that the conviction of the opposite truth forces itself upon one. It is obvious that for the purpose of a scientific knowledge of the formation of religious systems, we must not observe them in their fullest development, but rather turn to such of their brother-religions as have remained in a more stunted condition. Nor, again, should we deal, except very cautiously, with an extremely imaginative people, like the Greeks; for with them changes from any primitive form will be much more rapid and more complete than the changes in some more meagre systems. The fragmentary Teutonic myths, and the relics of these in mediæval superstition, are for this purpose sometimes more trustworthy than those of Greece; and partly on this account, partly because they are less familiar to the reader, we have drawn largely upon them for illustration in our chapters upon Aryan religion and Folk-tales. The most useful of all, however, is the religion of the Vedas, in so far as the Vedas give us an insight into the earliest faith of the people of India. Here we may often detect the etymology of a name which would be inexplicable if we only knew it in Greek or Latin and Norse. We have seen how this is the case in respect of the word Dyâus; and how the etymology of this word clearly shows, what from themselves we should never discover, that Zeus and Jupiter and Tyr are names which had originally the same meaning as a natural phenomenon. We say _originally_, because the Sanskrit is found by numberless examples (whereof we gave one, _duhitar_) to show an origin for many words whose origin is lost in other Aryan languages, and therefore to stand nearest to the primitive tongue of the Aryans. In this lies the whole force of the argument. If the old Aryans once used the same word for ‘heaven’ and for ‘god,’ it is impossible to believe that they had the power of separating at will the two ideas which we receive from these two words: for an examination of formal logic shows us that notions do not become completely distinguishable until they receive individual names. The inference is obvious that a considerable number, at any rate, of the gods of our Aryan ancestors were nature-gods in the strictest sense. It is equally true, however, that such divinities tend to fall into certain forms, and accommodate themselves to ideals which, or the germs of which, we may believe pre-existed in the human mind. It is thus that we have noticed the sun-gods and the heaven-gods fulfilling their separate functions, and answering to certain defined needs in the human heart.