22 BULLETIN 82, UNITED STATES NATIONAL MUSEUM VOLUME 1 



The piniuiles are not prismatic. With their short and broad segments witliout a 

 central constriction, of which tlic fii-st two are not broadened, they resemble most closely 

 the pinnules of certain of the Oligophreata. 



Completely plated disks occur only in the Oligophreata; the completely, though 

 thinly, plated disk of Aporometra therefore suggests reference to this group. The 

 spicules along the pinnule ambulacra are not distinctive. Very simUar spicules occur 

 both in the Macrophrcata (Antedon petasus, part 2, fig. 786, p. 366) and in the Oligo- 

 phreata iCyllometra manca [albopurpurea], part 2, fig. 777, p. 366). Such spicules are, 

 however, more frequently found in the Oligophreata than in the Macrophrcata. 



From this analysis it would appear that Aporometra finds its proper place in the 

 suborder Ohgophrcata. 



Confirmation of this allocation is seen in the general structm-e of the centrodorsal 

 with its 10 columns of cirrus sockets, but with the surface not divided into radial areas. 

 Centrodorsals of the same general natiu-e are seen in Asterometra macropoda (part 1, 

 figs. 189, 190, p. 235), Sliremeira arachnoides (part 1, fig. 205, p. 239), and Crinometra 

 concinna (part 1, fig. 206, p. 239). The centrodorsals of the few species in the Macro- 

 phrcata in which the cirri are arranged in 10 columns are quite different. 



Additional confirmation is seen in the chri which in general resemble those of Ptilo- 

 melra except that they are more or less flattened dorsoventrally at the tip. Broadening 

 of the cirri occurs elsewhere only in the family Colobometridae, though here the 

 broadening is usually confined to the pro.ximal portion. 



The relationships of Aporometra within the suborder Oligophreata remain to be 

 determined. Since the proximal pinmdes bear no comblike structure at the tip it 

 can not belong to the Comasterida, and since the pinnules are not prismatic it can not 

 be assigned to the Tropiometrida. The onlj- other superfamily in this suborder is the 

 Mariametrida. 



The superfamily Mariametrida corresponds roughly to the genus Himerometra as 

 it was understood in 1909 when Dr. H. L. Clark first described Himerometra paedophora. 

 So his reference of the new species to this genus, with the reservations that he gave, was 

 quite justified. 



At that time the families, genera, and species now grouped in the Mariametrida 

 were very imperfectly understood, and it was not at all unnatural that paedo2)hora 

 should have been considered as belonging to that assemblage. 



As it is understood at present the supcrfamdy Mariametrida includes the families 

 Zygometridae, Eudiocrinidae, Ilimerometridae, Mariametridae, and Colobometridae. 

 The entire absence of any suggestion of a dorsal crest on the oral pinnules, even on their 

 bases, and the more or less conical centrodorsal on which the cirri are arranged in ten 

 regidar columns, prevent the reference of Aporometra to any of these families. 



The structure of the centrodorsal in Aporometra resembles that of the centrodorsals 

 found in several genera and species in the families Asterometridae, Thalassometridae, 

 and Charitometridae of the superfamily Tropiometrida. The cirri also suggest those 

 in certain species of Asterometridae, Thalassometridae, and Ptilometridae, though 

 possibly they may more properly be compared with the cirri of some of the genera of 

 the Colobometridae which are, at least in the proximal portion, similarly broadened. 



The broadening of the cirri at the tip in Aporometra is, however, unique, as is also 

 the structiu-e of the articular faces of the radials, of the rosette, and of the very broad 

 and thin pinnide segments. 



