The fair complexioned Scandinavians, living in a cold country, and the bronzed Hindoos living in the torrid zone, has at first led to the supposition that the differences of coloration depended on temperature. But the Rohillas of Hindustan have a white skin, blue eyes, fair hair; whilst men with dark eyes and dark hair form the majority in certain districts of Ireland, Wales, and the Scottish highlands. The Gipsies, who came from India and spread over Europe since the 12th century, have in the cold countries, even to the Cheviot Hills, preserved the tawny complexion and the black eyes and hair of the Hindoos. The German colony in Paraguay, founded in the fifteenth century by the soldiers of Charles the Fifth, having remained pure by non-intermixture, presents a parallel instance of an Indo-European people remaining as fair under the tropic of Capricorn as on the banks of the Elbe.* Consequently, if there exist brown or fair IndoEuropeans, it does not depend on temperature. In reviewing the other climatic conditions, we find in the same way, by numerous examples, that they are incapable of producing the result in question.

We then arrive at influences of another order, which have not been considered as influencing the coloration, but which appeared to explain the variations in stature and muscular force. These are alimentation, mode of life, and subsidiarily industry, which leads to comfort. Two hypotheses are here face to face. It has been admitted that the social condition-that is to say, civilisation-by affording regular subsistence, abundant alimentation, gradually tends to increase both the height and physical force. Or, on the contrary, it has been admitted that civilisation, being unnatural, tends to develope the mind at the expense of the body, and in course of time renders man weaker and shorter. Both these hypotheses rest upon a certain number of facts, or rather coincidences. Thus the Græco-Latin people, civilised before the Germans, Scandinavians, and Slavonians, are shorter than the latter; but the Bas-Bretons are shorter than the Belgians, the Normans, and the Provençals, who have been civilised long before them. Many other facts might be cited in support of either of these hypotheses, which mutually destroy each other. It does not result from this that the conditions of existence have no influence upon stature; but thus much results, that the variation in stature of Indo-European peoples cannot be explained by these conditions.

It has finally been supposed that the variations of the cephalic

A similar Spanish family, remaining perfectly fair and purely Gothic, from non-intermarriage with darker types, and in position and rank holding for centuries official supremacy, is to be found in Yucatan at the present day.-ED. ANTH. REV.

type constituting brachycephaly and dolichocephaly might depend on intellectual culture; that the brain might, like the rest of the organs, be developed by exercise; that the most active organs of the brain might become more developed than the rest; and that consequently the degree and the nature of civilisation might modify both the volume and the form of the cranium. But, on the one hand, in taking a general view on the subject, we find that there subsists no relation between the brachycephalic and the dolichocephalic types as regards the intellectual value of races. The Teutonic races which occupy the first rank in the human series, are dolichocephalic, like the Ethiopian and Australian races, which stand last. Brachycephaly belongs to the Slavonians, the Turks, the Mantchoos, the Papuas, and numbers of other peoples occupying all degrees of the scale, without including the present French and South Germany, which are on the average nearly brachycephalic. On the other hand, from a special point of view of the Indo-Europeans, we find that the Scandinavians are more dolichocephalous than the Hindoos; these, again, are more so than the French; and that no relation can be established between the cephalic type of these different peoples and their civilisation either in the past or the present.

Having thus passed in review all these influences, and recognised that none of them explains the variations produced in the IndoEuropean nations, we arrive, by way of elimination, at another hypothesis which, without pretending to be rigorously demonstrated, possesses at least the advantage that it better explains all the facts and meets all objections. The existence of a primitive, or at all events an anterior population, at the arrival of the Indo-Europeans, has been established by human paleontology wherever persevering researches have been made. Among these aboriginal populations some were brachycephalic, others dolichocephalic; some were tall, others were short. The intermixture of the conquerors with the vanquished naturally explains all the variations in stature and cephalic type. Palæontology teaches us nothing as regards the colour of the eyes and hair of the aboriginals; but these have not altogether disappeared, and where they still exist, as in Hindustan and the region of the Pyrenees, they present, in reference to these characters, the greatest analogy to the Indo-Europeans adjoining them. All this is explained by the intermixture of races. The Indo-European idioms imported by the conquerors have prevailed over the indigenous languages, and have alone survived; but the indigenous people have not, for all that, disappeared. Although losing their language, their name, their nationality, they have not ceased to exist. The complete extermination of one race by another race is a phenomenon nearly impossible,

considering the condition in which the immigrant peoples found themselves. The intermixture of blood was an inevitable consequence of these immigrations, and, in proportion to the numerical preponderance of either race, the cross-breed resulting from this intermixture approached more or less to the indigenous or the foreign type. Nothing more is required in order to understand the diversity of the physical characters of the peoples now speaking the Indo-European idioms; and this theory of intermixture explaining all the facts, and being open to no objection, presents itself invested with every scientific probability.

This rather long digression seemed to us necessary, in order to show, by a very complicated example, how the analytical method may be applied to the study of anthropological questions.

The preceding example has, moreover, shown that great difficulties arise from the great diversity of intrinsic or extrinsic conditions amidst which the races under examination find themselves. All the characters they present are not of equal importance; some are more, others less, significative. If all tended in the same direction; if all the peoples of the same colour possessed the same stature, hair, cranial conformation—the same degree of intelligence, the same inclinations, the same language; if all of these were under the same, or at least a very similar climate; if all had at the same, or nearly the same, epochs arrived at the same social level—the task of anthropology would be an easy`one; but it is not so. The anatomical, physiological, psychological, climateric, and other facts combine and cross each other in a thousand ways. One character establishes an approximation, whilst other characters establish profound differences; and there result from this, continual contradictions, which, however, are and must be only apparent, and which will disappear when the whole truth becomes known, but which hitherto have given rise to difficulties and dissidence.

The naturalists also have had to contend with difficulties of the same nature, and if, after many failures, they have succeeded in giving to their science a positive character, it is because they have recognised the necessity of adopting a principle of co-ordination secured from the inroad of fancy. This principle is that of the subordination of character. This is not the place to expound it, to demonstrate its value, and to teach its applications. No one, moreover, ignores that it is

one of the most essential bases of the natural method.

The aim of the anthropologist should be to apply as much as possible to his science the principles of the natural method, a proposition which requires no demonstration. But among the characters diversifying the human group there are some which properly belong to it,

or are only found in a rudimentary state in other zoological groups. The particular considerations, according to which the naturalist establishes the subordination of characters, are thus insufficient for the anthropologist. In the presence of anatomical or morphological facts, whose relative value notably differs in the various degrees of the animal scale, there are other characters of quite a different order claiming a place which must be determined. When we consider that man is distinguished from other animals more by his intelligence than by his physical form, we can easily understand why some anthropologists have, in the classification of human races, assigned the first rank to psychical characters, by which humanity has acquired domination on the globe, and to assign the second place to physical characters, by which man so nearly approaches the anthropomorphous apes.

But the question should not be put in this way. Here it is not the question to distinguish the human genus from other groups, but to subdivide it into secondary groups as clearly defined and as natural as possible. It is necessary to base this division upon what is most fixed in the organisation of man, upon that which most resists the influences capable of modifying the individual or the race. Now, it is unquestionable that the physical characters are more permanent than the others, and that, consequently, they deserve the preference.

No doubt, languages, manners, industry, religion, all kinds of aptitudes, establish profound differences between the various races of mankind. But these characters, the study of which is as interesting as it is important, become frequently modified by circumstances, and may vary considerably in peoples of the same race. We cannot therefore assign to them any supremacy. There is, nevertheless, one which deserves special attention, and which plays a principal part in a great number of anthropological questions; so much so, that some others have felt justified in making it the almost exclusive basis for the classification of races. We speak of language. Linguistics render the most marked services to anthropology. Two peoples belonging to different races are separated from each other by several thousands of miles; they are so much strangers to each other that, neither in their respective histories, nor those of other peoples, is any mention found of their original parentage, and yet these peoples who have never heard of each other speak very similar idioms. The words are not the same, but the roots are. The grammar is nearly the same, and it becomes certain that these two languages had the same origin; consequently, the peoples speaking that language should, despite their actual dissemblances, have had common ancestors. On the other hand, there are two groups of races, which, since the origin of history, have always moved side by side, who have more than once intermixed,

who have interchanged their civilisation and religion, and who, as regards physical characters, present no very marked differences; such are the Indo-European and the Syro-Arab, improperly called Semitic, races. Now, despite their vicinity, their similarity of type, and various intermixtures, despite the fundamental community of facts, despite more or less durable political fusion, these two groups of races speak languages so distinct that the most eminent linguists have despaired of reducing them to a common origin. (See Renan, Histoire Générale et Système comparé des Langues Sémitiques. Paris, 1858.)

These inverse examples show the importance furnished by linguistics. These characters present, besides, a remarkable permanence. The spontaneous modifications introduced in the course of generations, either in grammar or words, however great they may appear, are of but a secondary order, the primitive type of the language continuing to subsist. This has been the case in all cases scientifically known; and these spontaneous modifications of words and grammatical forms constitute a sort of evolution subject to certain laws.

But

The linguistic characters have hereby acquired such a degree of precision that it has become easy to establish in languages methodical divisions and subdivisions, to distinguish a certain number of trunks dividing in primary, secondary, etc., branches, and thus to institute a taxonomy as regular, as positive, and as complete as that based upon physical characters. We may state here at once that in many cases the groups based on linguistics coincide pretty nearly with the groups based upon the anatomo-physiological study of human races. when these two orders of research lead to contradictory conclusions— as we have seen in the example of the variations of physical characters in peoples speaking the Indo-European tongues-it then becomes necessary to choose between the evidence of direct observation and that of linguistics, and to subordinate the characters drawn from language to those drawn from organisation, or vice versa. Naturalists readily give the preference to the latter; linguists place the former in the first rank. In order, then, to give to these difficulties a scientific solution, it becomes necessary to establish, on positive considerations, the relative value of the two orders of characters.

Every one admits that the distinctive characters of the races of man acquire value in proportion to their permanence. This is a general principle of natural history, and is also that of linguistics. The question, therefore, is whether the organisation of man is more or less permanent than his language. This question would not arise if the absolute and continuous immutability of the physical type were completely demonstrated. It is clear that languages become modified in