CONVERGENCE. I53 



branches of the Chahna, not the converse ; and the forms 

 of Reniera Hkewise merge in species not constant in 

 any character, and which the most careful observer is 

 unable to distinguish from the Chalinula. Therefore, if 

 the convergence or approximation of branches of various 

 origin cannot be rejected in the abstract, the most pro- 

 pitious case of coincidence is, nevertheless, limited to 

 the province of analogous formations, where, under like 

 conditions of adaptation, different families have been 

 driven to like expedients and differentiations, producing 

 complete similarity. A general survey of the organic 

 world likewise teaches us that in the higher regions this 

 overlapping of the ends of dissimilar parentage becomes 

 more and more incredible, and, as is shown by my study 

 on sponges, they can, in any case, occur only where the 

 organisms consist of very simple factors, highly variable 

 in a few directions, and very easily affected by external 

 conditions. 



When we referred above to the possibility of no 

 slight objection to the doctrine of Descent, we spoke 

 of another case of convergence. We mean, namely, 

 those similar final results in divergent series by which, 

 in highly organized groups of animals, of which the 

 reciprocal connection can be traced only through low 

 aboriginal forms, certain important organs exhibit the 

 greatest uniformiity of arrangement and integration. 

 It is, as yet, quite undecided where and when the true 

 insects separated themselves from the water-breathing 

 crabs ; nay, some naturalists incline to the opinion that 

 these two classes are derived from a more remote 

 common ancestor. Thus much is extremely probable, 

 tliat the severance between crab.'? and insects took place 



