240 PROCEEDINGS OP THE ACADEMY OP [1885. 



ingly small, without definite arrangement, and they cover the 

 surface radially and Lnterradlally. This is frequently the case in 

 the Silurian genera, Glyptocrinus, Periechocrinus, Melocrinus and 

 Reteocrinus. Their ventral covering resembles so closely the 

 <li-k of certain Comatulae, that it might appear as if this multi- 

 tude of irregular plates, which sometimes decrease in size toward 

 the periphery, and extend out to the free rays, could not be true 

 vault pieces. In proof, however, that this is the case, we refer to 

 Mr. St. John's carefully prepared diagram of Glyptocrinus ramu- 

 losus Billings, drawn from a specimen in the Canada Survey 

 Museum, and kindty loaned to us by Prof. Whiteaves. Of the 

 specimen only one-half of the calyx is preserved, and this is 

 imbedded in rock, exposing only the inner floor. Glyptocrinus 

 ramulosus is the largest species of the genus, and this facilitates 

 the study of the plates. Like all other species of Glyptocrinus 

 it has a large number of irregular interradial, interaxillary and 

 interbrachial plates, which meet* laterally over the arm openings, 

 and are continued to the summit, leaving no line of demarkation 

 between the plates of the two hemispheres. In the direction of 

 each arm opening the floor is distinctly grooved, and these 

 grooves or depressions, which diverge from the centre to the 

 arm furrows, evidently lodge the ambulacra. That the grooved 

 plates are not covering pieces, is shown by the fact that they 

 have the same irregular arrangement as the other plates. The 

 whole structure reminds us of Physetocrinus, and we have no 

 doubt that the vault in those two genera was built essentially on 

 the same plan. We find this further confirmed by the fact, that 

 in the Canada specimen the inner faces of the interradial and 

 interaxillary plates — but not any of the radial ones — are provided 

 with short nodes, such as are found in many Actinocrinidse, and 

 which serve there as pillars or partition walls between disk and 

 vault. 



In the Crotalocrinidae, which include Grotalocrinus and Enallo- 

 crinus, the whole ventral surface, in what appear to be the best- 

 preserved specimens, is composed of strong, convex plates, 

 without definite arrangement. In these specimens there is no 

 central piece, nor proximals, nor traces of ambulacra (Icon. Crin. 

 Suec, PI. 7, fig. 3 a ; PI. 8, figs. G, 7, and PI. 25, fig. 2) ; there are, 

 however, other figures of Angelin, apparently of a closely allied 

 species (Ibid., PI. 17, fig. 3 a), in which the plates paving the 



