214 



SCIENCE. 



[N. S. Vol. VII. No. 163. 



point there is a deeper kinship between an ex- 

 periment on the discrimination of space-inter- 

 vals and one on the discrimination of time-in- 

 tervals than there is between the latter and 

 experiments in simple reaction-time. And yet 

 in the author's arrangement the mere difference 

 in the object separates the various experiments 

 on discrimination by nearly the thickness of 

 the book, while the time elements bring reac- 

 tion experiments close to those on the estimate 

 of time-intervals, although the mental processes 

 investigated in these experiments are as differ- 

 ent as can be. 



But it is when interpreting experimental re- 

 sults that the author shows to least advantage. 

 If one were to generalize on the character of 

 the new movement in psychology from such 

 writings as this, one might say that the ' New 

 Psychology ' is wofully lacking in psychological 

 insight. There is tireless nicety in gathering 

 'facts,' only to make slovenly generalizations 

 which these facts do not warrant. Emerson 

 could have pointed to this as another illustra- 

 tion of his wide law of compensation. If the 

 older psychology was deficient on the side of 

 exact experiment, the new seems too often 

 wanting in any clear notion of what the experi- 

 ments prove. 



Many illustrations of this could be gathered 

 from the book, but the single instance of tap- 

 time must suffice. The rapidity with which 

 taps can be given on an electric key is assumed 

 by the author to indicate the rate at which we 

 can make separate acts of will. In truly scien- 

 tific work, however, it would seem appropriate 

 that the same exactness which is displayed in 

 recording and counting the taps should also be 

 used in determining whether these separate 

 movements of the finger are really due to 

 separate acts of will. To the present writer, at 

 least, the maximum rate of tapping seems to be 

 obtained by a peculiar muscular tension which 

 is preserved (it is true) by an act of will, but 

 the separate oscillations of the finger are no 

 more indicative of distinct acts of will than a 

 sustained rigidity would imply a separate voli- 

 tion for each unit of time the contraction was 

 maintained. Let us, by all means, have the 

 spark -method and full tables of mean variation 

 and all else that scientific accuracy may require. 



but let us not neglect the weightier matters of 

 the law. 



But, in spite of these and other defects, the 

 volume gives a really valuable account of the 

 more mechanical side of the experimental work, 

 and contains in small compass much that had 

 never been gathered into any single book. So 

 that Dr. Scripture has done good service in col- 

 lecting and arranging all this material. It is to 

 be regretted that the author's unfortunate 

 manner will, in too many cases, prevent even 

 his account of laboratory contrivances, in which 

 he is at his best, from getting the heartjf recog- 

 nition which the reader would otherwise be 

 sure to give. G. M. Stratton. 



University of California. 



Traits elementaire de mecanique chimique, fondSe 

 sur la Thermodynamique. By P. Duhem. 

 Paris, A. Hermann. 1898. Vol. II. Large 

 8vo. Pp. 378. Price, paper, 12 fr. 

 In treating the subject of chemical equilib- 

 rium one can classify the matter according to 

 components and subdivide according to vari- 

 ance, or one can reverse this, classifying ac- 

 cording to variance and subdividing according 

 to components. The first method is well 

 adapted to books on qualitative equilibrium, in 

 which the object is to get a clear view of the 

 behavior of a system as a whole. In books on 

 quantitative equilibrium it seems more rational 

 to group like equations together, and for that 

 reason it is better to discuss all nonvariant 

 systems and then all monovariant systems. 

 Since this second method has not yet been 

 adopted by any one, it is perhaps-not surprising 

 to find that Duhem has chosen the other in pref- 

 erence. The present volume, the second of the 

 series, treats of the laws describing one-com- 

 ponent systems and the systems which can be 

 made from these by addition or subtraction of 

 heat or work. This last statement may not be 

 clear without some explanatory comment. If 

 we start with solid ammonium chlorid we cer- 

 tainly have a one-component system, and this 

 is not altered by the fact that the vapor given 

 off by this substance is composed chiefly of am- 

 monia and hydrochloric acid. If we are nob 

 to make any distinction between a substance 

 which dissociates in the vapor phase and one 



