114 



weathered specimens, the two parts of the lancet plate are separate ; and 

 in many they appear to meet only the top and bottom of the cross section, 

 leaving a lozenge-shaped opening between them." (Geol. Mag., vol. ii, 

 p. 249,) In a large specimen of P. ohesus (Lyon and Cassiday which 

 was given to me by Mr. Lyon, a polished section shows that one of the 

 lancet plates is thus divided, but in general no trace of a suture can be 

 seen in these plates. 



There are several points in the structure of the ambulacra of Pentre- 

 mites that are well worthy of the study of those who have plenty of well 

 preserved specimens. Among these, I would direct special attention to 

 the markings in the ambulacrum of P. pyrifonnis. The median groove, 

 which I suppose to have been exclusively occupied by the ovarian tubes. 

 Bends off branches, right and left alternately, toward the sides of the 

 ambulacrum. These branches do not run directly to the ambulacrum 

 pores. Each of them terminates at a point between the inner extremities 

 ottwo of the pores- There is at this point a small pit which appears to 

 be the socket of an appendage quite distinct from the pinnule. The groove 

 does not reach the socket of the pinnule, which is situated further out, 

 between two of the pores. On the other hand a small groove runs from 

 each pore, inward, and terminates at another socket, about half-way 

 between the pore and the main median groove of the ambulacrum. It 

 would thus appear that besides the ordinary pinnules, there were two 

 other rows of appendages on each side of the median groove. 



The general conclusions at which I have arrived from the above, are, 

 that all the principal plates that compose the shell of Fentremites^ belong 

 to the perisomatic system of Professor Wyville Thompson ; that it is 

 doubtful whether or not the lancet plates are homologous with the radial 

 plates of the Crinoids ; and that the ambulacra are more complicated in 

 their structure than is generally supposed. 



7. On the Structure f the genus Nucleocrimis. 



Fig. 73. 



Fig. 74. 



Fig. 75. 



