200 



THE AMERICAN NATURALIST 



[Vol. LIII 



stimulus, as a growing plant toward the light, we might 

 seem to have an instance of such a directly purposive 

 action, i. e., the determination of the means by the end. 

 But several things must here be taken into consideration. 

 (1) It not infrequently happens that organisms are 

 drawn in an equally irresistible manner toward a fatal 

 stimulus, e. g., the moth to the flame; (2) w^e can not feel 

 sure, in every case, that the attainment of the goal is not 

 the outcome of random movements, unperceived by the 

 observer; (3) even where the response is indubitably 

 adaptive, and as direct and unfailing as a simple reflex, 

 it may be the outcome of a mechanism developed through 

 natural selection, i. e., the survival of random variations 

 w^hich were as frequently unadaptive as they were adap- 

 tive. The fact that some organisms still make suicidal 

 responses to less familiar stimuli favors this last view. 



Next, we may consider the phenomena of metabolism, 

 growth and development. We group these things to- 

 gether, because they can hardly be considered separately. 

 Growth is the outcome of metabolism, and development 

 of metabolism and growth. 



The phenomena revealed through studies of normal 

 physiology and embryology are obviously highly ''pur- 

 posive," in the sense that they have relation to the at- 

 tainment of an end, that end being the preservation of 

 the individual and the race. Nevertheless, they are be- 

 lieved by most biologists to be the outcome of a ''mech- 

 anism" the functioning of which presents no greater 

 difficulties, apart from complexity, than the working of 

 a clock or a steam-engine. 



When we come to consider the origin of this mech- 

 anism, we may mention three chief hypotheses, which 

 have been or still are held. (1) It may have been spe- 

 cially created by a super-mundane power in each indi- 

 vidual species of organism; (2) it may have gradually 

 developed out of simple beginnings by the "selection" 

 or survival of random variations which were as likely to 

 be unadaptive as adaptive; or (3) it may have gradually 



