language of the warlike invader or peaceful immigrant may yield, in a few generations, to the tongue of the mass of the population, as the Northman s was replaced by French, and modern German gives way to English in the United States. Judging, then, by the extirpation and adoption of languages within the range of history, it is obvious that to classify mankind into races, Aryan, Semitic, Turanian, Polynesian, Kafir, &c., on the mere evidence of language, is an intrinsically unsound method. From the earliest times in which nations have been classified by languages,
its unrestricted use has vitiated sound ethnology.Nevertheless, under proper restrictions, speech affords information as to the affinities of races only second in value to that derived from physical characteristics. As a rule, language at least proves some proportion of ancestry. It could hardly happen that one people should come into so close a relation to another as to supplant its language, without strong intermixture of race in the next genera tion. This is true in the extreme case of the West Indian coloured population, among whom the majority are now crossed with European blood, so that in each succeeding generation the proportion of absolutely pure Negro families becomes less. Still more fully is it true of coloured races in Mexico or Brazil, whose Spanish or Portuguese language represents at least a large European element of ancestry. Thus in India many millions of people, whose blood is predominantly that of the darker indigenous race, never theless speak dialects of the languages of the fairer Aryans ; but then they are for the most part distinctly mixed races of partly Aryan ancestry. With these facts before us, it is not difficult to determine the principles on which the ethnologist may use language as partial evidence of race. In the first place, it strengthens the evidence of bodily characters. Thus in South Africa the Zulu seems by colour, features, shape of skull, &c., to be, if not an abso lute Negro, of a mixed and modified Negro type. This view of his origin is strengthened by the fact that the Zulu language belongs to the peculiar prefixing family which extends so widely among the Negro nations farther north. So the Hottentot language, in its evident connec tion with that of the Bushmen, adds its weight to the physical argument, that these two are descendants more or less mixed and varied from a single race, small, yellow, crisp-haired, and speaking an inflectional monosyllabic language, articulated with clicks. In the second place, language may prove race-connection where bodily character istics, though they do not contradict, do not suffice. Thus, comparing the dark Andalusian with the fair Swede, we ask the question, whether there is distinguishable common parentage between these two varieties of the white man 1 The anatomist might hesitate here. Nor, indeed, is the physical problem nearly solved, but at least a partial solution is involved in the philologist s proof that the two peoples speak languages inherited at some remote period from a common Aryan tongue, and must therefore have had a common element in their ancestry of at least suffi cient strength to carry language with it. Thus each linguistic family affords at least partial evidence of race, proving, for instance, the existence of a common ancestry of the Irishman and the Russian, of the Jew and the Maltese, of the Tahitian and the Malagasy, though in such pairs of races the actual amount of common ancestry may be less than that of the different race-elements with which it has combined.
As regards political nationality and the history of civilisation, the evidence of speech is of still greater weight. In many cases of the mixture of nations the language of the dominant civilisation prevails, as where Latin dialects superseded the native tongues in Western Europe, and Germanic languages encroached on Turanian in Finland, on Slavonic in Russia, and on Keltic in the Scotch High lands. In other cases, where one nation has received elements of civilisation from another, language is apt to keep record of the process by adopting foreign words and ideas together. Thus the language of the barbarian Turks has absorbed masses of Arabic, which itself had in like manner absorbed Persian, when Persia was the fountain- head of early Moslem culture. In the same manner Dravidian languages of South India have been saturated with words and phrases from Sanskrit and its related dialects, so that a page of Tamil literature is of itself the proof of a non- Aryan race having received from an Aryan race a whole system of religion, philosophy, and social order. The most extreme cases of such verbal indication of foreign influence are to be found in languages of low races of America and the Pacific, which have adopted from European languages not only terms for imported arts and ideas, but names of such numerals as G and 7, pre viously expressed by more clumsy native combinations. Thus the language of any people, though less effective than was once believed as a means of determining its place in the classified order of mankind, does, to some extent; indicate its physical, and, to a still greater extent, its intellectual ancestry.
at the lowest and highest known levels of culture are separated by a vast interval; but this interval is so nearly filled by known intermediate stages, that the line of continuity between the lowest savagery and the highest civilisation is unbroken at any critical point. The Australians and forest Indians of Brazil may be taken as the lowest modern savages whose thought and life have been investigated with any thoroughness ; while other less accurately-studied tribes are in some respects inferior even to these. An examination of the details of savage life shows not only that there is an immeasurable difference between the rudest man and the highest lower animal, but also that the least cultured savages have themselves advanced far beyond the lowest intellectual and moral state at which human tribes can be conceived as capable of existing, when placed under favourable circumstances of warm climate, abundant food, and security from too severe destructive influences. In fact, the Australian or Brazilian savage has already attained to rudimentary stages in many of the characteristic functions of civilised life. His lan guage, expressing thoughts by conventional articulate sounds, is the same in essential principle as the most cultivated philosophic dialect, only less exact and copious. His weapons, tools, and other appliances, such as the hammer, hatchet, spear, knife, awl, thread, net, canoe, &c. T are the evident rudimentary analogues of what still remains in use among Europeans. His structures, such as the hut, fence, stockade, earthwork, &c., may be poor and clumsy, but they are of the same nature as our own. In the simple arts of broiling and roasting meat, the use of hides and furs for covering, the plaiting of mats and baskets, the devices of hunting, trapping, and fishing, the pleasure taken in personal ornament, the touches of artistic decora tion on objects of daily use, the savage differs in degree but not in kind from the civilised man. The domestic and social affections, the kindly care of the young and the old, some acknowledgment of marital and parental obliga tion, the duty of mutual defence in the tribe, the authority of the elders, and general respect to traditional custom as the regulator of life and duty, are more or less well marked in every savage tribe which is not disorganised and falling to pieces. Lastly, there is usually to be discerned amongst such lower races a belief in unseen powers pervading the universe, this belief shaping itself into an animistic or
spiritualistic theology, mostly resulting in some kind of