92 THE CRINOIDEA CAMERATA OF NORTH AMERICA. 



plates by the letters '' rx'' in tlie accompanying diagrams.* There now re- 

 mained among the plates of the proximal ring but four which could possibly 

 be taken for the orals ; and this led us to inquire whether the central plate 

 alone might not be a coalesced representative of the five orals of recent forms. 



From internal casts we observed that this plate occupies the centre of 

 radiation, and that not only the ambulacra, but also the nerve cords, meet 

 beneath it. It was this structure principally which led us to the assumption 

 that the central plate represented the five orals collectively, and that the 

 four large proximals, and two smaller ones, were interradial " vault " plates, 

 corresponding with the first interradials of the abactinal side. This seemed 

 to us the more probable, as in the dorsal cup a division of the first interra- 

 dial into two halves by an anal plate is a frequent occurrence among Palseo- 

 zoic Crinoids. It also seemed to explain why in Haplocrinm and allied forms, 

 in which there is no anal plate, the central piece seemed to be surrounded 

 by five plates instead of six, supposing, as before stated, that Haplocriniis had 

 a small oral surrounded by five interradial plates, and AUagecrinus, Coccocrinus, 

 and CiiUcocrinus five interradials, but no orals. In this we differed from Gotte, 

 Carpenter, Zittel, and Neumayr, who all agreed that the 8cheitelplatten were 

 the orals. t 



This was the state of the question in 1888, when we came into posses- 

 sion of a very large number of fine specimens of Haplocrinus mespiliformis 

 in various stages of preservation, and found to our astonishment that 

 such a thing as a " central " plate does not exist in the genus. We now 

 saw that the ventral disk consists of but ^ye large plates; that we had 

 mistaken a mere fracture for a suture; and that the part which we sup- 

 posed to be a separate piece was a tongue-like prolongation of the posterior 

 plate, projecting in between the other four plates, and sometimes surmounted 

 by a small node (Plate III., Fig. 12^). This discovery left no room for doubt 

 that the large ventral plates of Haplocrimis, and of the Laviformia generally, 

 actually represent the five plates composing the unopened oral pyramid of 

 the Pentacrinoid larva before it moved away from the radials, as had been 

 contended by Carpenter and Goette. 



So long as a central plate was recognized in Haplocrinus, we saw good 

 reason to believe in the existence of a similar plate in other groups of the 



* Revision, Part HI. Plate YH. Pigs. 2, 3, 4, 5, 6, 8, 9, 10, and Plate VHI. Pigs. 1 and 2. 



t Our theory of the relations of the summit plates, in conformity with these views, was discussed in the 

 Revision, Part III. pp. 44 to 59, and afterwards in greater detail in our paper on the Summit Plates, in the 

 Proceedings of the Academy of Natural Sciences, Philadelphia, March 29, 1887. 



