842 



SCIENCE 



[N. S. Vol. XL. No. 1041 



that the second law of thermodynamics does 

 not always apply to life-processes is certainly 

 inadequate; there is exact experimental evi- 

 dence that the first law (that of conservation) 

 holds for organisms; and the storing of solar 

 energy by chlorophyll is in no sense evidence 

 that the second law is evaded. There seems 

 in fact to be a fundamental misconception in 

 this part of the author's argument. He holds 

 that life may play the part of the Maxwellian 

 demon under appropriate circumstances (page 

 118), and defends this view on the ground that 

 the laws of molecular physios are statistical 

 in their nature and might be different if it 

 were possible to control the movements of indi- 

 vidual molecules; such control, it is implied, 

 is possible to the vital entelechy. It seems 

 to the reviewer, however, that the application 

 of the second law to gases or solutions implies 

 simply a tendency of the freely moving mol- 

 ecules to uniform distribution; the resulting 

 homogeneity can be prevented only by adding 

 energy to, or abstracting it from, part of the 

 system; even Maxwell's demon has to work a 

 partition which resists the impact of the faster 

 molecules — a consideration which shows that 

 any coordination or sorting of molecules would 

 in itself involve the performance of work. 

 Johnstone's supposition, however, is that the 

 vital entelechy can, without altering the total 

 energy of the system, control or direct the 

 otherwise uncoordinated motions of the indi- 

 vidual molecules; and that the purposive or 

 directed character of the individual organism's 

 life, and also of the whole organic or evolu- 

 tionary process, is conditional on the existence 

 of such an agency, and is indeed the character- 

 istic expression of its activity. He thus main- 

 tains, in effect, that physiological processes 

 are unintelligible unless we can assume the 

 existence of some such directive agency pecu- 

 liar to life, which can vary the nature, inten- 

 sity and direction of the physico-chemical proc- 

 esses and coordinate them in the interest of the 

 organism. This " entelechy " is what imparts 

 their distinctive quality to life-phenomena. 



It has long seemed to the reviewer that fail- 

 ures or deficiencies in the physiological anal- 

 ysis of complex or delicately adjusted functions 



form no sufficient ground for rejecting such 

 methods of investigation as in their nature 

 inadequate. Vitalists, however, are fond of 

 this kind of attack; and both Haldane and 

 Johnstone adduce instances which they believe 

 make it incredible that physico-chemical proc- 

 esses, unguided by an entelechy, could ever 

 form the basis of vitality. At present our 

 knowledge of the physiology of embryonic 

 development and of certain types of form- 

 regulation is especially defective; and such phe- 

 nomena are cited more frequently than any 

 others as proving the inadequacy of physico- 

 chemical analysis. Driesch's " logical proof of 

 vitalism," quoted in the present book, is an 

 instance of this tendency; even relatively 

 simple processes like muscular contraction and 

 nerve conduction remain largely mysterious, 

 and we find also scepticism as to the possibility 

 of any satisfactory account of these processes 

 in physico-chemical terms (c/. page 100 of the 

 present book). 



A twofold reply to this type of vitalistic 

 argument may be given. First, it is to be 

 noted that the failure of physico-chemical anal- 

 ysis is often due to mere complexity of condi- 

 tion. But complexity, as such, does not 

 introduce any essentially new problems; it 

 simply makes more difficult, and may for a 

 time make impossible, the task of analysis. 

 Provided that the more elementary processes 

 forming a complex process are characterized 

 by constancy in their nature and in the condi- 

 tions of their occurrence, any degree of com- 

 plexity in the total process is possible. Ordi- 

 nary experience with complex artificial sys- 

 tems, of a mechanical or other kind, verifies 

 this contention; we find that there is no limit 

 other than that set by practical expediency to 

 the complexity of a system whose component 

 parts operate and interact in a constant 

 manner. In all such cases smaller and simpler 

 parts are taken as units from which higher 

 compound units are built up, and these second- 

 ary units are then similarly utilized for the 

 construction of more complex systems; these 

 may be still further combined, and so on. The 

 one indispensable condition is that there should 

 be an essential invariability in the operation 



