78 DIVERGENCE AND PARALLELISM 



some Annelid worms, e.g., Syllidae, strobilation is 

 superadded to the normal segmentation. As I 

 have indicated in the diagram (Fig. 2), the 

 Appendiculata are here regarded as a parallel 

 stem to the Vertebrata, having a common 

 Coelenterate origin. This scheme is in accord- 

 ance with the conviction that the somites of 

 Annelids and Arthropods have been evolved 

 along their own line of descent, independently 

 of the somites of Vertebrata. 



The metamerism of Appendiculata and Verte- 

 brata is therefore, on this view, a conspicuous 

 example of convergence, and as such I regard 

 it, taking into account the facts which have been 

 made known concerning it. If this view is 

 correct, as I believe it to be, there can be no 

 question of comparing the regional differentia- 

 tion of Arthropods, such as Limulus and the 

 Scorpion, with that of Vertebrata. The whole 

 comparison, except by way of a possible con- 

 vergence here and there, is ruled out of court. 

 The existence of convergent strobilation is in 

 itself presumptive evidence of the possibility of 

 convergent metamerism, and there are other facts 

 which render it extremely probable and, to my 

 mind, practically certain. 



Instances of parallel convergence are so 

 numerous and so common that we begin to 

 realise that convergence is a regular and not 

 an exceptional phenomenon. The most striking 



