has been a favourite idea among anthropologists that the savage races inhabiting parts of India give us a correct idea of what the Aryans of India were before they were civilized. It may safely be said of this as of other mere ideas, that it may be true, but that there is no evidence to show that it is true. At all events it takes much for granted, and neglects, as it would seem, the very lessons which the theory of evolution has taught us. It is the nature of evolution to be continuous, and not to proceed _per saltum_. Therein lies the beauty of genealogical evolution that we can recognize the fibres which connect the upper strata with the lower, till we strike the lowest, or at least that which contains what seem to be the seeds and germs of early thoughts, words, and acts. We can trace the most modern forms of language back to Sanskrit, or rather to that postulated linguistic stratum of which Sanskrit formed the most prominent representative, just as we can trace the French _Dieu_ back to Latin _Deus_ and Sanskrit _Devas_, the brilliant beings behind the phenomena of nature; and again behind them, _Dyaus_, the brilliant sky, the Greek _Zeus_, the Roman _Iovis_ and _Iuppiter_, the most natural of all the Aryan gods of nature. This is real evolution, a real causal nexus between the present and the past. It used to be called history or pragmatic history, whether we take history in the sense of the description of evolution, or in that of evolution itself. History has generally to begin with the present, to go back to the past, and to point out the palpable steps by which the past became again and again the present. Evolution, on the contrary, prefers to begin with the distant past, to postulate formations, even if they have left no traces, and to speak of those almost imperceptible changes by which the postulated past became the perceptible present, as not only necessary, but as real. Perhaps the difference is of no importance, but the historical method seems certainly the more accurate, and the more satisfactory from a purely scientific point of view.