THE IXNER TISSUES OF ANIMALS. 351 



tiou most along the tract where they are most abundant ; will, 

 as before, simultaneously tend to form new molecules of their 

 own type ; will, as before, make the line along which they lie 

 one of easier transfer for the molecular agitation. Every 

 repetition will help to increase, to integrate, to define more 

 completely, the course of the escaping molecular motion- 

 extending its remoter part while it makes its nearer part 

 more permeable will help, that is, to form a line of discharge, 

 a line for conducting impressions, a nerve. 



Such seems to me a not unfair series of deductions from 

 the known habitudes of colloids in general and the organic 

 colloids in particular. And I think that the implied nature 

 and properties of nerve, correspond better with the observed 

 phenomena than do the nature and properties implied by 

 other hypotheses. Of course the speculation as it here stands 

 is but tentative, and leaves much unexplained. It gives no 

 obvious reply to the questions what causes the formation of 

 nerves along some lines rather than others ? what determines 

 their appropriate connexions ? questions, however, to which, 

 when we come to deal with physiological integration, we may 

 find not unsatisfactory answers. Moreover it says nothing 

 about the genesis of ganglia. A ganglion, it is clear, must 

 consist of a colloidal matter equally unstable, or still more 

 unstable, which, when disturbed, falls into some different 

 molecular arrangement, perhaps chemically simpler, and gives 

 out in so doing a large amount of molecular motion serves 

 as a reservoir of molecular motion which may be suddenly 

 discharged along an efferent nerve or nerves, when excite- 

 ment of an afferent nerve has disengaged it. How such 

 a structure as this results, the hypothesis does not show 

 But admitting these shortcomings it may still be held that 

 we are, in the way pointed out, enabled to form an idea of 

 the actions by which nervous tissue is differentiated. 



303. A speculation akin to, and continuous with, the laat^ 

 is suggested by an inquiry into the origin of muscular tissue 

 55 



