The Crinoids FEo:\r Dr. S. Iîock's I^Ixtetmtiox to Japan IHN. 7)7) 



Fiirtlier it ma\- be ti)<ii tliese possibly existing small species 

 partly make liylu'ids between themselves and perhaps partly also form 

 crossings with pi'oximal species in the genera Gnmaste)\ Cornaniheria 

 and GomanfJim. If ihcse hybrids are fertile they will after a couple 

 of generations be split into countless different types. Supposing, for 

 instance, that a lorin without cirri, with 111 and IV Br-s : 4. <nid with 

 a Comantims-comh (restricted to proxinuil [linnules) were crossed with 

 a form with cirri, with III and ]\ IJr-s : 1'. and with Comaslrr-comh 

 (that is short, high combs dispersed to distal pinnules). Fnrthei-, sup- 

 posing that such a hybrid were fertile and therefore in the 2'^ gene- 

 ration was normally split in (he different possibilities of coml)itui(i((n. 

 Then one would get foi-ms that might be referred to all the genera 

 known within the sub-fam. GomaderiiKr. (If, e. g., the first 8 gens 

 as heterozygotic might be represented with different forms: rudimen- 

 tary cirri, mixture of 2 and 4 components in the III and IV division- 

 seri^, one would get 10<S different possibilities). The great variability 

 in the division-series even in the same specimen seems to be a proof 

 of the probability of the hetorozygotical nature of certain specimens of 

 C. parricirra. 



In a special ])aiicr I will treat a 3'' possible influencing cause, 

 viz. alteration in the mode of catching food. 



Though in many cases the Comas(e)'i(h show primitive charac- 

 eristics such as clumsy formation, frequently (in the genus Comaster) 

 a piimitive way of arm-divisioi]. a number of components in the divi- 

 sion-seiies that is ^ery indefinite, it is evident that in other charac- 

 teristics the family is in fast development and has entered quite a 

 specific direction of evolution. A lot of the characteristics which are 

 engaged in such a development have this apparently caused by the 

 changed manner of catching food and arc treated in connection with 

 this question. 



The \aiiability of the cirii is probably caused by another 

 factor and their regression, \\hich is observable in different species of 

 the Comasterid family, I suppose to be caused by the dorsal hooks on 

 the distal pinnules having taken over the anchoring function that 

 otherwise falls upon the cirri. These dorsal hooks are to be found in 

 most of the Comatulids, though usually visible only by strong magni- 



