686 THE AMERICAN NATURALIST [Vol. XLIII 



why the octopus could not be assumed to have been de- 

 scended from stalked ancestors along the same lines. It 

 seems to me that the echinoderms are rather like the 

 bivalve molluscs or the Crustacea which contain both free 

 and fixed types, the former in the great majority. The 

 crinoids are the only recent fixed echinoderms ; but in the 

 fossil crinoids, as Lang pointed one in Marsupites and I 

 independently showed in Uintacrinus, there are forms 

 which exhibit no evidence of ever having been attached; 

 in fact the evidence is quite the other way. In these 

 forms the centrale may be, instead of the centro-dorsal, 

 really the dorso-central, in which case we should get an 

 interesting homology with the echinoids. 



The association of the holothurians with the echinoids, 

 and hence with the crinoid-echinoid stem, seems to me to 

 be abundantly justified. The following classification of 

 the echinoderm groups is proposed as showing the inter- 

 relations of these groups better than any of the synopses 

 previously published. 



Phylum ECHINODERMATA 

 I. Subphylum Echinoderhata Heteroradiata. 

 A. Pelmatozoa. 



1. Crinoidea. 



B. Ovozoa. 



1. Eehinoidea. 



C. Vermiformes. 



1. Holothuroidea (Bohadschoidea). 

 II. Subphylum Echinodermata Astroradiata. 



A. Ophiobrachiata. 



1. Ophiuroidea. 



B. Stellarides. 



In this table the sequence of groups must not be taken 

 to represent a phylogenetic line ; in no class of animals is 

 a phylogenetic sequence more difficult of conception than 

 in the echinoderms. While the heteroradiate echino- 

 derms are, judged by ordinary standards, more perfect 

 than the astroradiate, judged from the echinoderm point 

 of view solely, they can not be considered so well de- 

 veloped as the latter. 



