208 DEGENERATION sec. 



important, formerly diagnostic, ordinal character a new form 

 of generic value has been produced." 



The causes of the disappearance of the four-rayed spicules 

 in 0. Schmidt's opinion remain in obscurity. But he also 

 holds that their presence or absence in Caminus does not 

 depend on selection, since other spicules, uniaxial and stellate, 

 are present, and because the genus in no respect gives the 

 impression of degeneration, ratlier seems " to stand at the 

 acme of life." Schmidt classes these spicules among those 

 characters which, as he justly says, are somewhat indefinitely 

 called " morphological," that is, immaterial, indifferent. 



Indeed, in Caminus the uniaxial spicules are beginning 

 here and there to degenerate, but independently of the 

 degeneration of the quadriaxial. 



It may of course be objected that the quadriaxial spicules in 

 Caminus were once useful, and that they are now disappearing 

 in consequence of the cessation of selection, because other use- 

 ful characters have appeared in this genus of sponges wliich 

 make the presence of the quadriaxial spicules unnecessary. 

 But proof of this is wanting. 



And the extreme variation of the skeletal parts of sponges 

 in one and the same species indicates in the clearest way 

 that the form of these parts is a character whose variation, 

 at all events within very wide limits, remains entirely un- 

 affected by selection. 



Oscar Schmidt points out further that numerous other 

 cases in sponges have been described by Haeckel and himself, 

 in which the organisms are beginning to change into new 

 species by the disappearance of certain forms of their skeletal 

 structures. And I am able to add that in the markings of 

 animals — e.g. butterflies — characters everywhere degenerate 

 whose present or former use cannot be discerned, which we 

 must resjard as non-essential. 



Weismann supposes that even in those cases in which 



