THE SYSTEM OF PIIYLLOPHORIN.E. 



67 



specimens often prevents accurate counting. Much more useful in the 

 systematic seems to be the distribution of pedicels and papilla?, and 

 above all, I should lay great emphasis upon the ordinal rank of the 

 tentacular canals, which may be said to be fairly constant. 



6. The system, of PhyUophorinœ. 



Semper's classification of Dendrochirotae into three subfamilies, 

 Stichopoda, Gastropoda and Sporadipoda, based on the distribution of 

 pedicels [42, PP- 47, 61 and 64], as well as Bell's [5, p. 254] and 

 Lampert's [18, p. 114] proposal to divide the group into Dccachirotae 

 and Polychirota: according to the number of tentacles, were both 

 rejected by LUDWIG [26, P- 321 and 323] as being in no harmony 

 with the natural relationship of the genera. 



ÖSTERGREN iu his recent paper [37, P- 212] has offered a new 

 view on the system of Holothurioidea, establishing under the family 

 Cucumariidae the three subfamilies, Cucumariinae, Phyllophorinae and 

 Psolinae. These three subfamilies are practically synonymous with 

 Decachirotae, Polychirotae and Gastropoda respectively. Nevertheless, 

 OsTERGREN's above view seems to me to be acceptable, since, firstly, 

 it is desirable to derive family or subfamily name from the type 

 genus ; secondly, since the polychirote-gastropodous form Thcelia 

 LUDWIG can not find its proper position in any of the old sub- 

 families, and thirdly, since the Cucumariinae and Phyllophorinae differ 

 from each other not only in tentacle number but also in other 

 anatomical features. 



Tabulating the various forms of the Phyllophorinae with respect 

 to the arrangement of ambulacral appendages, we have the 

 following : — 



