interesting social history of France, and is additionally instructive in showing how the skill of the statistician comes to the aid of historical erudition in the attempt to construct a picture of past social conditions.

Of greater interest to the general reader is the first part of the book, the Introduction sur la Statistique. It is extremely moderate in tone, the author not claiming for statistics even the name of a science, except for that part dealing with population, which he desires to call "démographie." At the same time it contains capital observations on the purpose of statistical inquiry, the value of the results, statistical method and graphic illustrations, with a concise history of statistics. I commend it to those who desire an intelligible and lucid statement of the scientific criteria by which we are to direct statistical inquiry and measure the value of the results. It will temper the zeal of those over-ardent statisticians who stand ready to take the statistics of anything, from the morality of a community down to the condition of its city sewers; while it will perhaps open the eyes of those scoffers who assert that as good statistics can be obtained for one side of any question as for the other. Statistical inquiry is a method of scientific investigation adapted to certain purposes. To apply the method to unsuitable purposes is as absurd as to call in the jeweller to mend your iron gate, or to send your watch to the blacksmith for repair. The true value of statistics will appear only when we comprehend their limitations.

M. Levasseur points out how necessary it is to arrange inquiries so as not to excite the passions or offend the prejudices of the people from whom information is desired. So late as 1841 there were riots in the city of Toulouse on the taking of the census, the report having got abroad that it was to be used for fiscal purposes. In France, questions in regard to religious belief are apt to arouse sectarian passions, where the old religious differences have not yet been forgotten. Again, a census may easily be led into inquiries which are not worth the expense of making; as for instance in regard to the color of the hair and eyes, which, although sometimes of ethnological interest, is generally valueless on account of the uncertainty of the classification and the mixture of races. Again, some statistical facts are so purely administrative in character that they gather themselves, so to speak; as the convictions for crime, or the imports and exports. Others are not administrative, but can easily be gathered by administrative officers, such as births, deaths and marriages. Still others are neither administrative nor readily gathered by administrative officers, as special facts in regard to agriculture, factories, wages, indebtedness. It is much easier to bring the results of the first two classes to some degree of perfection than those of the last, and the statistics of the former are correspondingly more trustworthy. Finally, M. Levasseur points out how by skilful arrangement and com

bination a few inquiries may lead to a great variety of information. These few illustrations will suffice to show the knowledge and acumen of the author.

R. M. S.

Darwinism and Politics. By D. G. RITCHIE, M.A. London, Swan Sonnenschein & Co., 1889. — 101 pp.

It is impossible for a great idea to become dominant in any department of science without affecting men's thoughts on all other subjects that keenly interest them. This is illustrated in our own time by the ideas of the "struggle for existence" and "the survival of the fittest"; from biology they have been inevitably transferred to sociology and economics. Ernst Haeckel used them, justifiably enough perhaps, to prove the impossibility of realizing the ideal of equality which he attributed to socialists, and at the same time to demonstrate, as against Virchow, how harmless Darwinism was to society. Mr. Herbert Spencer and his followers go very much further than this, and regard the two principles as an unassailable basis for a policy of laissez faire, and as a justification for denouncing the very moderate amount of restraint upon the individual that English radicals have of late years been advocating. Fortunately the growth of historical culture has been too great to permit Mr. Spencer's sociology to take much hold upon students of economics. But among the professed students of biology, there is not infrequently found a disposition to pooh-pooh efforts at social reform as unscientific; and "the struggle for existence" has already become a phrase in the mouth of the man of the world.

Mr. Ritchie's little essay is a timely criticism of this application of biological conceptions to social problems. He begins by calling our attention to the ambiguity in the term "fittest," and to the fact that the question is much more complex than is often supposed, in that the struggle is not only between individuals, but also between groups in the same nation and between different nations. Then, coming to the demand that natural forces should be allowed free play, he points out the difficulty of distinguishing between "natural" and "artificial" without falling back into the unhistorical conception of a state of nature; and he justly observes that the not infrequent confession on the part of evolutionists, that the mitigation of the primitive struggle has led to a moral advance, concedes the desirability of displacing "natural" by human and conscious selection. There is in truth a struggle of ideas as well as of physical forces, and they have an equal claim to a fair chance. The recognition of this fact leads us to the fundamental difference. between social evolution and all other development, namely, the appearance of consciousness. And yet, conscious effort, though we are often

obliged to have recourse to it in order to counteract the unrestrained play of "natural" forces, is not necessarily in antagonism to them, and may frequently operate only to accelerate or facilitate their action. Mr. Ritchie strengthens his general argument by specific illustrations, and concludes by showing how desirable it is that conscious effort should take the shape of positive legal institutions, partly because they are necessary to give effect to opinion, partly because of their educative value. The educational influence of institutions is usually overlooked by those evolutionists who magnify the force of heredity. The discussion that is now going on as to whether acquired characteristics are ever transmitted by descent tends, as Mr. Ritchie remarks, to weaken the claim of heredity as the preponderating element in the formation of character. W. J. ASHLEY.

The Primitive Family. By C. N. STARCKE, Ph.D., of the University of Copenhagen. New York, D. Appleton & Co., 1889. -315 PP.

Die Indogermanischen Verwandtschaftsnamen. Ein Beitrag zur vergleichenden Alterthumskunde. Von BERTHOLD DELBRÜCK. [Abhandlungen der philologisch-historischen Classe der Kgl. Sächs. Gesellschaft der Wissenschaften. Bd. xi, No. 5.] Leipzig, S. Hirzel, 1889.-228 pp. In publishing an English translation of Dr. Starcke's Die primitive Familie in ihrer Entstehung und Entwickelung (Leipzig, 1888), the Appleton's have rendered a service to English and American students. In the first place, the book gives an extensive collection of facts and a useful review of theories. Besides this, it contains much suggestive criticism and some original points of view. Its field is wider than its title it treats not merely of the family, but of kinship and of the organization of the clan and tribe.

From Dr. Starcke's point of view (as from Morgan's) it is not the family but the clan which determines primitive kinship. The bond of the clan is community of blood; the family is based upon a mixture of blood. Whether kinship shall be traced through the male line or through the female depends, therefore, rather upon the organization of the clan than upon that of the family. If the wife enters the husband's clan, we shall find agnatic kinship; if the husband joins the wife's, we shall find Mutterrecht. The prevalent and persistent tendency is in the former direction, towards father-right; and father-right seems to develop somewhat as follows. Clans tend to become exogamous because, according to Dr. Starcke, the man who captures a wife from a neighboring clan avoids all interference with the established family relations

of his own clan. It might be added that it is cheaper to steal a wife than to buy one; and that in primitive society capture is the most respectable means of acquiring property. With wife-capture arises the idea of property-right in the wife, and in the children born of the wife; and this property-right is the basis of father-right and of agnatic kinship. In marriage by purchase, the basis of the husband's and father's right is the same. This, I take it, is what Dr. Starcke means by saying that fatherhood does not rest originally on the fact of procreation, but is essentially "juridical" (pages 121-127); but I am not altogether sure that I understand his line of reasoning. In other passages (e.g. on page 37) he seems to distinguish a primitive paternal authority based on pure force, and a later father-right, of a legal character, not based on force at all. But he may mean only that in the development of paternal authority, as in the development of all property rights, there is an earliest stage in which the social sanction has not yet converted force into right.

Society also tends, from the earliest period, to monogamy. Polygamy can never be general, because the number of women is never greatly in excess of the number of men. Polyandry Dr. Starcke regards as sporadic, rejecting McLennan's theories. He also rejects the theory that there is any necessary connection between the levirate or the niyoga and polyandry. So far his views coincide with those of most writers. But Dr. Starcke goes further: he rejects entirely the theory of primitive hetairism or promiscuity, and the derivation of mother-right from the impossibility of determining fatherhood. He disputes the priority of mother-right. Explaining mother-right on "residential" grounds, he tries to show that a community originally agnatistic may change its habits and derive relationship entirely through the female line. In a cattlebreeding community, for instance, men make it their first object to increase the number of stock. In such a community the head of the family will sell his daughter as early and for as high a price as possible. In the agricultural community, on the other hand, the chief demand is for labor. In such a community, the head of the house will not only oppose the departure of his daughter; he will seek to induce her wooer to become one of his household. Hence the herding community will recognize agnatistic kinship only, while mother-right will appear with the development of agriculture (pages 99, 100). This argument seems to me singularly fallacious. It bases the probability of wife-purchase entirely on the willingness of the woman's father to sell. But it takes two to make a bargain; and it may be insisted, with more reason, that the development of wife-purchase will depend on the desire of the wooer to buy. If this point of view be accepted, Dr. Starcke's data afford a conclusion directly opposed to that which he reaches. The herding community will maintain mother-right, because the man does not care

to purchase a wife; but in the agricultural community, where labor is the prime requisite, the man who cannot steal a wife from a neighboring and hostile community will buy one within the tribe as soon as he is able. Dr. Starcke's whole attack upon the theory of primitive mother-right seems to me inconclusive. He unquestionably shows that mother-right may exist, on "residential" grounds, when fatherhood is not uncertain. But surely an institution which has arisen on one ground may persist on another. He unquestionably shows that father-right may exist on “juridical" grounds (ie. as a property right), when the customs of the community are such as to render paternity anything but certain; and he insists that if father-right does not rest primarily upon procreation, mother-right can never have rested upon the uncertainty of the father. This does not seem to me a necessary inference. He further endeavors to show that the couvade - which may be best described as the "confinement" of the husband - does not point to primitive mother-right or represent an attempt to find an analogous basis for father-right (pages 51, 52). Here again Dr. Starcke gives reasons which might account for the persistence of the institution, but which seem inadequate to explain its origin. His statement that the Arowaks and Macusis, who practise the couvade, have the system of exclusive maternal kinship, is not borne out by the data he furnishes on pages 37, 38.

All Dr. Starcke's conclusions on this point are conditioned by certain premises which are hardly likely to command general acceptance. According to his canons, we should never seek the origin of a custom in primitive conditions if its existence is explicable under contemporaneous conditions. He is disinclined to reason back from symbols to earlier realities and from fictions to primitive facts.

Dr. Delbrück's study of Indo-Germanic Kinship Names is a work of narrower scope than Dr. Starcke's, but certainly of no less value. The bulk of the treatise is purely philological, but in the last forty-five pages (Sachlicher Theil) the author sums up his conclusions. The ancient Aryan system, he thinks, is strictly agnatic. In the oldest literature of India the father's brother plays a more important part than the mother's brother. There is no philological evidence that polyandry was customary. There is no philological evidence of primitive promiscuity and resultant mother-right. On the other hand, the philological evidence does not, and cannot, disprove the mother-right theory. If the anthropological arguments in favor of that theory be deemed conclusive, all that the philologist can say is that the Indo-Germanic peoples seem to have emerged from promiscuity and mother-right before that period in their development with which Indo-Germanic philology deals. Dr. Delbrück's conclusions are largely identical with those of Dr. Edward Washburn Hopkins, whose treatise on the Ruling Castes in Ancient India he