1885.] NATURAL SCIENCES OF PHILADELPHIA. 289 



it. The two latter groups possess a series of large anal plates, 

 arranged horizontally, and these form a proboscis with a furrow 

 at its ventral side. This proboscis was incorporated into the sac 

 in a somewhat similar manner as the lower arm joints and 

 pinnules into the disk of the Neocrinoidea. In this structure the 

 two groups have close analogies with the recent genus Thaumato- 

 crinus. In that genus, however, the row of anal plates does not 

 enter the perisome, but forms an independent solid appendage in 

 the shape of a cone, which apparently has no functions, as the 

 anal opening is persomic, and we regard this peculiar appendage 

 as a remarkable instance of atavism. 



It is probable that in the latter Poteriocrinidae and Encrinidse, 

 the interradials and summit plates became finally resorbed, and 

 the perisome was more or less restricted to the ventral disk, as 

 in these genera the sac dwindled down to a small conical tube, 

 which probably disappeared in Encrinus before reaching maturity. 



That the openings along the ventral sac are not genital open- 

 ings, as suggested lately by Trautschold, need not be discussed, 

 as most of the Fistulata have well-developed pinnules, and these 

 are not prehensile organs as supposed by him, but are continua- 

 tions of the arms which contained the genital glands. Neither 

 is it true that the ventral sac is frequently present or absent in 

 the same species. It existed in every individual, but is rarely 

 preserved in the fossil, and is often obscured by the arms. 



Nothing is known from actual observation of the perisome of 

 the Ichthyocrinidffi, and little if anything of the construction of 

 the ventral side in any of their genera. The interradial plates 

 of the dorsal side have been described by us as movable, some- 

 what irregular in form and arrangement, and upon this, prin- 

 cipally, we based our conclusion that the plates of the ventral 

 covering were movable, in some cases perhaps squamous. In 

 Onychocrinus onl} r there has been observed by Lyon and us 

 indistinct traces of a ventral covering, but too imperfect to give 

 much information either as to the real nature of the plates, or 

 as to their arrangement. Carpenter regards all interradials of 

 the dorsal side as cahvx plates, and all those succeeding them 

 and located ventrally as parts of the disk. We admit that the 

 latter may have a superficial resemblance to the small, irregular 

 and movable perisomic plates of Extracrinus and other Neo- 

 crinoidea to which he alludes. But we do not understand why a 

 20 



