appreciable change has occurred in three or four thousand years, we should be obliged to assume a far more remote date for the first branching off of such races from a common stock than the supposed period of the Aryan migrations, and the dispersion of that language over many and distant countries.
But Mr. Crawfurd has, I think, himself helped us to remove this stumbling-block, by admitting that a nation speaking a language allied to the Sanscrit (the oldest of the eight tongues alluded to), once probably inhabited that region situated to the north-west of India, which within the period of authentic history has poured out its conquering hordes over a great extent of Western Asia and Eastern Europe. The same people, he says, may have acted the same part in the long, dark night which preceded the dawn of tradition.[1] These conquerors may have been few in number when compared to the populations which they subdued. In such cases the new settlers, although reckoned by tens of thousands, might merge in a few centuries into the millions of subjects which they ruled. It is an acknowledged fact, that the colour and features of the Negro or European are entirely lost in the fourth generation, provided that no fresh infusion of one or other of the two races takes place. The distinctive physical features, therefore, of the Aryan conquerors might soon wear out and be lost in those of the nations they overran; yet many of the words, and, what is more in point, some of the grammatical forms of their language, might be retained by the masses which they had governed for centuries, these masses continuing to preserve the same features of race which had distinguished them long before the Aryan invasions.
There can be no question that if we could trace back any set of cognate languages now existing to some common point
