He gives another and much longer example, which the reader may be spared, adding that there are altogether about 200 particles, as many as ten of which may be piled up on any given stem. The process also often involves great phonetic changes, by which the original form of the elements becomes disguised, as, for instance, in the English _hap'oth_ = half-penny-worth. The attempt to determine the number of words that might be formed in this way on a single stem, such as _igdlo_, a house, had to be given up after getting as far as the compound _igdlorssualiortugssarsiumavoq_ = he wants to find one who will build a large house. It is clear that such a linguistic evolution implies both the postulated isolation from other influences, which must have disturbed and broken up the cumbrous process, and also the postulated long period of time to develop and consolidate the system throughout the New World. But time is still more imperiously demanded by the vast number of stock languages, many already extinct, many still current all over the continent, all of which differ profoundly in their vocabulary, often also in their phonesis, and in fact have nothing in common except this extraordinary polysynthetic groove in which they are cast. There are probably about 75 stock languages in North America, of which 58 occur north of Mexico. But even that conveys but a faint idea of the astonishing diversity of speech prevailing in this truly linguistic Babel. J. W. Powell[766] points out that the practically distinct idioms are far more numerous than might be inferred even from such a large number of mother tongues. Thus, in the Algonquian[767] linguistic family he tells us there are about 40, no one of which could be understood by a people speaking another; in Athapascan from 30 to 40; in Siouan over 20; and in Shoshonian a still greater number[768]. The greatest linguistic diversity in a relatively small area is found in the state of California, where, according to Powell's classification, 22 distinct stocks of languages are spoken. R. B. Dixon and A. L. Kroeber[769] show however that these fall into three morphological groups which are also characterised by certain cultural features. It is the same, or perhaps even worse, in Central and in South America, where the linguistic confusion is so great that no complete classification of the native tongues seems possible. Clements R. Markham in the third edition of his exhaustive list of the Amazonian tribes[770] has no less than 1087 entries. He concludes that these may be referred to 485 distinct tribes in all the periods, since the days of Acuna (1639). Deducting some 111 as extinct or nearly so, the total amounts to "323 at the outside" (p. 135). But for such linguistic differences, large numbers of these groups would be quite indistinguishable from each other, so great is the prevailing similarity in physical appearance and usages in many districts. Thus Ehrenreich tells us that, "despite their