208 COSMIC PHILOSOPHY. [pt. i. 



a still more striking instance, let us remember how Adam 

 Smith's elucidation of the principle of " division of labour," 

 in sociology, suggested to Goethe the conception of a " division 

 of labour " in biology, and thus heralded Von Baer's magni- 

 ficent discovery that organic development is a progressive 

 change from homogeneity to heterogeneity of structure. And 

 let us note how this discovery in biology has lately reacted 

 upon all preceding departments of investigation, strengthening 

 the nebular theory in astronomy and the theory of the pro- 

 gressionists in geology ; and thus ultimately reacting upon 

 our philosophy by giving us, for the first time, a scientific 

 doctrine of the evolution of the physical universe. 



Enough has been alleged to prove that the Comtean view 

 of the progress of science fails to account for more than a 

 limited portion of the facts of that progress. Instead of the 

 sciences aiding each other, with few and unimportant ex- 

 ceptions, only in the hierarchical order in which Comte has 

 placed them, we perceive that they have continually been 

 aiding each other in all directions at once. The more 

 complex sciences have all along been assisting the simpler 

 ones, and these have often been delayed in their progress for 

 want of the assistance which the former have ultimately 

 furnished. There has, therefore, been no such thing as a 

 progressive evolution of the sciences in a linear order ; but 

 there has been a consentaneous evolution, in which the 

 advance of each science has been a necessary condition of 

 the advance of all the others. 



It thus appears that Comte unduly simplified the problem. 

 His classification well enough expresses the order of develop- 

 ment of the sciences, in so far as their development has 

 depended merely on the relative simplicity or complexity of 

 the phenomena with which they have had to deal. It rests 

 upon the assumption that, with few and unimportant ex- 

 ceptions, the progress of generalization has been from the 

 simple to the complex. Now this is not the case. The 



