Thus， if I am right， the same ancient pair of deities was variously known among the Greek and Italian peoples as Zeus and Dione， Jupiter and Juno， or Dianus (Janus) and Diana (Jana)， the names of the divinities being identical in substance， though varying in form with the dialect of the particular tribe which worshipped them. At first， when the peoples dwelt near each other， the difference between the deities would be hardly more than one of name； in other words， it would be almost purely dialectical. But the gradual dispersion of the tribes， and their consequent isolation from each other， would favour the growth of divergent modes of conceiving and worshipping the gods whom they had carried with them from their old home， so that in time discrepancies of myth and ritual would tend to spring up and thereby to convert a nominal into a real distinction between the divinities. Accordingly when， with the slow progress of culture， the long period of barbarism and separation was passing away， and the rising political power of a single strong community had begun to draw or hammer its weaker neighbours into a nation， the confluent peoples would throw their gods， like their dialects， into a common stock； and thus it might come about that the same ancient deities， which their forefathers had worshipped together before the dispersion， would now be so disguised by the accumulated effect of dialectical and religious divergencies that their original identity might fail to be recognised， and they would take their places side by side as independent divinities in the national pantheon.
This duplication of deities， the result of the final fusion of kindred tribes who had long lived apart， would account for the appearance of Janus beside Jupiter， and of Diana or Jana beside Juno in the Roman religion. At least this appears to be a more probable theory than the opinion， which has found favour with some modern scholars， that Janus was originally nothing but the god of doors. That a deity of his dignity and importance， whom the Romans revered as a god of gods and the father of his people， should have started in life as a humble， though doubtless respectable， doorkeeper appears very unlikely. So lofty an end hardly consorts with so lowly a beginning. It is more probable that the door (janua) got its name from Janus than that he got his name from it. This view is strengthened by a consideration of the word janua itself. The regular word for door is the same in all the languages of the Aryan family from India to Ireland. It is dur in Sanscrit， thura in Greek， tür in German， door in English， dorus in old Irish， and foris in Latin. Yet besides this ordinary name for door， which the Latins shared with all their Aryan brethren， they had also the name janua， to which there is no corresponding term in any Indo-European speech. The word has the appearance of being an adjectival form derived from the noun Janus. I conjecture that it may have been customary to set up an image or symbol of Janus at the principal door of the house in order to place the entrance under the protection of the great god. A door thus guarded might be known as a janua foris， that is， a Januan door， and the phrase might in time be abridged into janua， the noun foris being understood but not expressed.
From this to the use of janua to designate a door in general， whether guarded by an image of Janus or not， would be an easy and natural transition.