PHYSIOLOGICAL INTEGRATION IN ANIMALS. 381 



out the higher stages of nervous development, can be seen 

 only still more vaguely. Nevertheless, it is comprehensible 

 that as functions become further divided, there will arise the 

 need for sub-connexions along which there may take place 

 secondary equilibrations subordinate to the main ones. It is 

 manifest, too, that whereas the differentiation of functions 

 proceeds, not necessarily by division into two, but often by 

 division into several, and usually in such ways as not to leave 

 any two functions that are just complementary to one an- 

 other, the restorations of equilibrium cannot be so simple as 

 above supposed. And especially when we bear in mind that 

 many differentiated functions, as those of the senses, cannot 

 be held complementary to any other functions in particular; 

 it becomes manifest that the equilibrations that have to be 

 made in an organism of much heterogeneity, are extremely 

 complex, and do not take place between each organ and some 

 other, but between each organ and all the others. The pecu- 

 liarity of the molecular motion propagated from each organ, 

 has to be neutralized by some counter-peculiarity in the 

 average of the molecular motions with which it is brought 

 into relation. All the variously-modified molecular motions 

 from the various parts, must have their pluses and minuses 

 mutually cancelled: if not locally, then at some centre to 

 which each unbalanced motion travels until it meets with 

 some opposite unbalanced motion to destroy it. Still, involved 

 as these actions must become, it is possible to see how the 

 general principle illustrated by the simple case above sup- 

 posed, will continue to hold. For always the molecular 

 motion proceeding from any one differentiated part, will 

 travel most readily towards that place where a molecular 

 motion most complementary to it in kind exists — no matter 

 whether this complementary molecular motion be that pro- 

 ceeding from any one other organ, or the resultant of the 

 molecular motions proceeding from many other organs. So 

 that the tendency will be for each channel of communication 

 or nerve, to unite itself with some centre or ganglion, where it 



