The fascination of this particular branch of philological research is apparent when we recollect that in this case, in the case of these remote Eastern ancestors of ours, philology is almost the only window through which we can look out on them. In most subsequent periods of history we have many other ways, besides the study of language, of discovering the outward circumstances of men’s lives. Historical records, archaeology, ethnology, folk-lore, art, literature, all come to our help in considering, say, the ancient Egyptian civilization; but it is not so with the Aryans. Here ethnology and archaeology tell us practically nothing, anthropology a little, and the rest nothing at all. If we wish to cross the darkness which separates us from this period we must lay down a little plank of words and step delicately over it. And in such romantic circumstances it is hardly surprising that we should find a veritable army of scholars and philosophers, both professional and amateur, jostling each other upon that plank with such vigour that the bridge and its burden have often seemed in danger of vanishing quietly together into the abyss. The central principle upon which philologists have worked is this, that if a word occurs to-day in a fair sprinkling of the Aryan languages, then that word existed in the Aryan parent-language, and therefore the thing of which it is the label existed in some form or other in the primitive Aryan civilization. Conversely, if an object or an idea is found to have a different name in most of the Aryan languages, it was sometimes assumed that that object was not known to the Aryans before their dispersion. But this negative deduction soon came to be regarded as unsafe, and there are indeed many reasons why the whole method is limited and uncertain. For instance, even in one language it is constantly happening that when a new thing or a new idea comes into the consciousness of the community, it is described, not by a new word, but by the name of the pre-existing object which most closely resembles it. This is inevitable. We have to proceed from the known to the unknown in language as in life; but language lags behind life and words change more slowly than things or ideas. When railways first came in, their rolling-stock consisted of a string of vehicles resembling the old horse coach so exactly that it was said later that “the ghost of a horse stalked in front of the engine”. Although this is no longer the case, we still call these vehicles _carriages_ or _coaches_, and look like continuing to do so. To take an even more patent example, when a modern Englishman or American uses the very old Celtic word _car_, we all know what he means: yet it would be an error to deduce from this that the principle of internal combustion was known in pre-Christian times in Wales, Ireland, Cornwall, Brittany, and probably Rome (Latin ‘carrus’, a cognate word). Moreover, we can see at once that the