THE OUTER AND INNER TISSUES OF ANIMALS. 305 



wise chemically altered. Immediate metamorphoses of this 

 kind are often obviously produced in ova by changes of their 

 media. At the outset, therefore, before yet there existed 

 any such differentiation as that which now usually arises by 

 inheritance, these environing agencies must have tended to 

 originate a protective envelope. For a modification produced 

 by them on the superficial part of the protoplasm, must 

 either have been a decomposition or else the formation of a 

 compound which remained stable under their subsequent 

 action. There would be generated an outer layer of substance 

 that was so molecularly immobile as to be incapable of further 

 metamorphoses, while it would shield the contained proto- 

 plasm from that too-great action of external forces which, by 

 rapidly changing the unstable equilibrium of its molecules 

 into a relatively stable equilibrium, would arrest development. 

 Evidently organic evolution, whether individual or general, 

 must always and everywhere have been subordinate to these 

 physical necessities. Though natural selection, beginning 

 with minute portions of protoplasm, must all along have 

 tended to establish a molecular composition apt to undergo 

 this differentiation of surface from centre to the most favour- 

 able extent, yet it must all along have done so while con- 

 trolled by this process of direct equilibration. 



Contrariwise, the many and great unlikenesses among the 

 dermal structures of creatures inhabiting the same element, 

 cannot be ascribed to any such cause. The contrasts between 

 naked and shelled Gastropods; between marine Worms and 

 Crustaceans, between soft-skinned. Fishes and Fishes in 

 armour like the Pterichthys, must have been produced entirely 

 by natural selection. Environing forces are, as before, the 

 ultimate causes; but the forces are now not so nluch those 

 exercised by the medium as those exercised by the other in- 

 habitants of the medium; and they do not act by modifying 

 the surface of the individual, but by killing off individuals 

 whose surfaces are least fitted to the requirements: thus 

 slowly affecting the species. Still the dermal skeleton bristling 

 6(? 



