NATURi: AND CAUSE O^ VARIATION. 299 



isms, however, the conditions are immensely more complicated than in the 

 mathematical or physical sciences, and it will necessarily be a long time before 

 we can even begin in one single instance to follow the order of events in any 

 given variation in organisms. 



In the explanation of the origin of variation in organisms the only assump- 

 tion we need make is that the original unit of organic matter was possessed 

 of the attributes which characterize organic matter to-day — motion, sensa- 

 tion, growth, and reproduction. This assumption can not meet with any 

 serious objection unless we change our ideas and definition of organic units. 

 Granted the existence of one single organic unit endowed as above, there is 

 no reason for introducing further complications by the explanation of phe- 

 nomena through undemonstrable hypotheses, because the fact of variation in 

 organic units can be explained solely through their existence in a natural 

 world surrounded by varying conditions of existence. The original organic 

 unit could not do otherwise than encounter changing conditions, or do other- 

 wise than react to the stimuli it encountered, even though it were so lowly 

 organized that all stimuli were perceived as one and reacted to by the sole 

 response of contraction. In this we have all the conditions requisite for 

 variation in full conformity with the law of trial and error. 



The simplest existing organisms, greatly specialized, and the heredity 

 product of countless generations of similar simple organisms, react to stimuli 

 according to the method of trial and error. Thus Jennings says, regarding 

 the responses of the lowest organisms to stimuli : 



In these creatures [Amoebae] the behavior is not, as a rule, on the tropism plan — a 

 set, forced method of reacting to each particular agent — but takes place in a much more 

 flexible, less directly machine-like way, by the method of trial and error. This method 

 involves many of the fundamental qualities which we find in the behavior of higher 

 animals, yet with the simplest possible basis in ways of action; a great portion of the 

 behavior consisting often of but one or two definite movements — movements that are 

 stereotyped when considered by themselves, but not stereotyped in their relation to the 

 environment. 



It may be objected that Jennings is dealing entirely with sensory charac- 

 ters, reactions; yet wherein lies the real objection? Are not activities, 

 responses, behavior, indissolubly associated with structures ? Should we look 

 for one law of structure and another of activities? Undoubtedly structures 

 respond to stimuli. We know it to be a fact, and they follow in structural 

 variations the same law of trial and error. Were this not so, the observed 

 variations would not fall so truly into the probability of error distribution. 



There does not exist, as far as I can discover, any real reason why organic 

 and inorganic variations are not the manifestation of the universal law of 

 error (trial and error) ; nor is there any necessity of introducing latencies or 

 predeterminations into the explanation of organic variations. Theoretically, at 



