268 THE PRINCIPLES OF HEREDITY 



432. We have just seen that the performance of any com- 

 plex action becomes more easy by frequent repetition, till, 

 if the action be repeated frequently enough, the performance 

 of it becomes automatic. In common but erroneous parlance 

 it becomes instinctive. Mr. Lewes supposed that some of 

 this acquired facility in performance is bequeathed by succes- 

 sive parents to successive offspring, whereby it is accumu- 

 lated during generations to such an extent that the remote 

 descendant has inborn the facility which the remote ancestor 

 acquired only with effort and difficulty ; which is the same 

 thing as saying that the character that arose in the ancestor 

 under one form of stimulation, arises in the descendant 

 under quite another. In this manner, according to Lewes, 

 do actions at first intelligent and accompanied by a sense 

 of mental effort become mechanical and instinctive, passing 

 first through a border-space where they are neither quite 

 intelligent nor altogether instinctive, but partake of the 

 nature of both. 



433. But setting aside for the moment the conclusion we 



useful i.e. the most habitual actions. Thus we may see in a general 

 way how such nervous machinery may at last come to be differentiated 

 into specially distributed anatomical structures, which, on account of 

 their special distribution, are adapted to minister only to particular co- 

 ordinations of muscular movements. That is to say, we are thus able 

 to understand the rise and development of Reflex Action." (Mental 

 Evolution in Animals, pp. 30-3.) 



Mr. Spencer's theory appears plausible at first sight. Examined 

 closely it is found to consist of a tissue of vague assumptions, many of 

 which are opposed to known facts. A mass of undifferentiated proto- 

 plasm is assumed. Such a mass exists nowhere in nature except perhaps 

 among the lowest unicellular organisms in whom nerves never arise. 

 It is assumed that stimuli applied to a given point will at first be 

 diffused through the mass, but that, presently, channels (i. e. nerves) 

 will be established through which the stimuli will pass to definite 

 points. The very vague analogy of flowing water is given. But water 

 does not wear out channels in a perfectly level and homogeneous surface. 

 Without initial differentiation in the protoplasm the stimuli would 

 continue to be diffused through the mass. It is lightly assumed that 

 the changes which stimuli cause in one individual (the parent) will 

 reappear in quite another individual (the offspring) in the absence of the 

 stimuli. Lastly, the power of developing under the influence of use, 

 which is possessed only by certain of the structures of the highest 

 animals, is supposed to be a property common to all protoplasm. The 

 fact that actions tend to become automatic in the highest animals is 

 no evidence that channels for the discharge of stimuli are readily 

 established by use in all nervous tissue. It is evidence only that certain 

 channels are readily established in a, particular class of nervous tissue 

 which a very prolonged course of evolution has adapted to that special 

 function. 



