REVIEIVS 701
much more summary than that of Ross. Tarde commits himself to theses, in the direct line of his argument, that obviously call for closer analysis. For instance, in general his often forced appeal to his stock explanation "imitation." Our objection to his use of this "principle" is that he has long ago reasoned himself away from his earlier and better-known judgments about its role among social forces. He has described the companion facts of "opposition," of " invention," and of "adaptation." Yet he seems unable to restrain himself from implying that "imitation" is the same pass-key to all social changes which he claimed that it was before he had reached these later perceptions. Again, in particular, M. Tarde gets dangerously distant from his base of supplies in such theorems as that legitimate authorities rest chiefly upon delief, tyrannical authorities upon desire (36-44 et passim). In the latter case the author is partially aware of his own omissions, to be sure, for he acknowledges that the terms "legitimate" and "illegiti- mate" involve a certain vicious dualism. In the same connection his insistence that the family is the source of all political power is followed (35) by admission that the family is, at first, shop, church, school, regiment, state ; and that the nature of power differs according to the one of these sources {sic) which predominates. By this admission Tarde throws his own argument into confusion about the very dis- tinction between source and channel of power with which he sets out (23). On the whole, the first part of the book seems to emphasize demand for a treatment of the genesis of political authority on a new scale of thoroughness.
In the second part Tarde opens up a mass of material for kinetic theory with originality of conception quite as distinct as that shown in Ross' program. Indeed, in chap. 10 Tarde throws down the gauntlet in this fashion(i87) : "It is our business to formulate laws of causation, of logical causation, which at once state and partially exemplify rules proclaimed by the philosophers of history, and at the same time exhibit frequent exceptions to the rules. Moreover, these logical laws will lead to other solutions both more comprehensive and more penetrat- ing." In showing how he would go to work to carry out this plan, Tarde suggests problems enough to convince the most skeptical that there is work to do in this field. Perhaps the same skeptics are not yet in a state of mind to be convinced that the work is worth doing, but this may be too much to expect at this early date. For instance, the author points out (192) that " Social evolution is a problem in which it is important first of all to distinguish with care the independent variables
