266 THE PRINCIPLES OF HEREDITY 



from one another. They supposed, therefore, that there 

 were border spaces where they merged. Thus " Spencer 

 regards instinct as compound reflex action and the precursor 

 of intelligence, while Lewes regards it as lapsed intelligence, 

 and therefore necessarily the successor of intelligence. Thus, 

 while Lewes maintains that all instincts must originally have 

 been intelligent, Spencer maintains that no instinct need 

 ever have been intelligent." l Professor Romanes, from whom 

 I have quoted, is in partial agreement and disagreement 

 with both Mr. Lewes and Mr. Spencer, thinking that in 

 some cases the one is right and in some cases the other. 2 



1 Mental Evolution in Animals, p. 256. 



8 Romanes gives a condensed but clear account of Mr. Spencer's 

 theory of the origin of reflex action and its concomitant nervous tissue. 

 " The following is the theory whereby Mr. Herbert Spencer seeks to 

 answer this question, and in order fully to understand it we must begin 

 by noticing the effects of stimulation upon undifferentiated protoplasm. 

 A stimulus, then, applied to homogeneous protoplasm, which is every- 

 where contractile and nowhere presents nerves, has the effect of giving 

 rise to a visible wave of contraction, which spreads in all directions from 

 the seat of stimulation as from a centre. A nerve, on the other hand, 

 conducts a stimulus without undergoing any contraction, or change of 

 shape. Nerves, then, are functionally distinguished from undifferentiated 

 protoplasm by the property of conducting invisible or molecular waves 

 of stimulation from one part of an organism to another, and so establish- 

 ing physiological continuity between such parts without the necessary 

 passage of visible waves of contraction. 



" Now, beginning with the case of undifferentiated protoplasm, Mr. 

 Spencer starts from the fact that every portion of the colloidal mass is 

 equally excitable and equally contractile. But soon after protoplasm 

 begins to assume definite shapes, recognized by us as specific forms of 

 life, some of its parts are habitually exposed to the action of forces 

 different from those to which other of its parts are exposed. Conse- 

 quently, as protoplasm continues to assume more and more varied 

 forms, in some cases it must happen that parts thus peculiarly situated 

 with reference to external forces will be more frequently stimulated to 

 contract than are other parts of the mass. Now in such cases the 

 relative frequency with which waves of stimulation radiate from the 

 more exposed parts, will probably have the effect of creating a sort of 

 polar arrangement of the protoplasmic molecules lying in the line 

 through which these waves pass, and for other reasons also will tend 

 ever more and more to convert these lines into passages offering less 

 and less resistance to the flow of such molecular waves i. e. waves of 

 stimulation as distinguished from waves of contraction. And lastly, 

 when lines offering a comparatively low resistance to the passage of 

 molecular impulses have thus been organically established, they must 

 then continue to grow more and more definite by constant use, until 

 eventually they become the habitual channels of communication between 

 the parts of the contractile mass through which they pass. Thus, for 

 instance, if such a line has been established between the points A and B 

 of a contractile mass of protoplasm, when a stimulus falls upon A, a 

 molecular wave of stimulation will course through that line to B, so 



