1879.] NATURAL SCIENCES OF PHILADELPHIA. 289 



from Dendrocrinus only in having the suture between the sections 

 of the compound plate sloping instead of horizontal, that by this 

 in itself trifling alteration, which required no modification in the 

 form or construction of other plates, the lower portion of ^ie 

 original compound radial became transformed into an anal plate. 

 This was the first step towards a Poterocrinus anal arrangement, 

 and in fact to complete it required only the interposition of a 

 third small plate (PL 16, Fig. 7). In confirmation of this idea it is 

 very significant that Dendrocrinus is essentially a Lower Silurian 

 genus, that Homowinus is restricted to the Niagara group, and 

 Poteriocrinus is pre-eminently a subcarboniferous form. 



The step from Dendrocrinus to Cyathocrinus (PL 16, Fig. 8) 

 was equally simple, and required only the consolidation of the 

 compound plate into one, the simple anal plate being already 

 developed in the former. Sometimes, however, in genera wherein 

 otherwise the Cyathocrinus anal arrangement prevails, there is 

 found alongside of the single anal plate in some species occa- 

 sionally only in a few specimens, and in Barycrinus and Botryo- 

 crinus almost as a rule a supplementary anal piece obliquely in- 

 terposed toward the right side, standing as a witness to the com- 

 mon origin of this and all anal plates. That there has never been 

 observed a single instance in which the anal area was directed 

 toward the left not even an abnormal case is most significant, 

 and is strong evidence in favor of our opinion, that the ventral 

 sac originated iir the right posterior ray. It proves also that the 

 modifications which we have mentioned as taking place in the 

 family in geological succession, are occasionally found within the 

 limits of a genus. 



It is to be regretted that the ventral sac, owing to its position, 

 hidden between the arms, is so rarely observed. A better know- 

 ledge of this organ, we have no doubt, would enable us to base 

 upon it excellent generic distinctions, and it might perhaps assist 

 in establishing subdivisions in this family. Very little is known, 

 for instance, of the distribution and position of the pores and 

 fissures, of the anal opening, etc., and nothing as to its internal 

 organization. 



In the Lower Silurian the ventral sr,c attained large dimensions. 

 In the age of the Niagara group, Upper Silurian, it became re- 

 markable for its singular form, in some genera coiling upon itself, 

 or bending in all directions. In the Devonian and earlier Subcar- 



