1885.] NATURAL SCIENCES OP PHILADELPHIA. 271 



plates, nor with one another, and each summit plate has a totally 

 isolated position. In the smaller species and younger specimens, 

 however, all summit plates are connected, showing that those 

 small inserted plates result from excessive growth, and are intro- 

 duced to increase the capacity of the visceral cavitj'. 



Higher orders of summit radials exist in comparatively few 

 genera. We must admit that the descriptions which we gave of 

 these plates, although correct as to certain species, cannot he 

 applied to the Palseocrinoidea generally, nor even to all Actino- 

 crinidae. Manjr of them have but a single radial, and the plates 

 which we took to be radials in most of them, prove to be inter- 

 axillaries and interbrachials, which often attain a larger size than 

 the surrounding plates. A very conspicuous case of this kind is 

 Borycrinus, in which the large spiniferous plate above each ray 

 is not a primar} r radial as we had supposed, but an interaxillary, 

 for the bifurcation of the ambulacral tube takes place beneath 

 the preceding plate. The misconception of these plates in this 

 and other genera led us to suppose that the arrangement of some 

 of the summit plates was more or less disturbed in all species 

 with a large number of arms, while in fact we had searched for 

 plates which are unrepresented. The arrangement of the summit 

 plates, as a rule, is very regular, and only disturbed by the anal 

 tube. They are readily recognized even in Megistocrinus , Stroto- 

 crinus and Teleiocrinus as seen by our diagrams (PL 8, figs. 1, 

 3, 5, and PL 4, fig. 4). 



It has been stated that the proximals, in all cases in which 

 they have been recognized by us, consist of more than five plates, 

 generally of seven, and we have asserted, which has been 

 accepted by Carpenter (Chall. Rep., p. 167), that the two outer 

 plates at the azygous side are equivalent to, and take the place 

 of a fifth large one, being separated from each other by anal 

 plates or the proboscis. The structure is well shown by our 

 diagrams, but in examining them it must be borne in mind that 

 the plates marked Xr are radials, and not interradials as hereto- 

 fore supposed. The more central the position of the anal 

 aperture, and the larger the size of the tube, the greater is the 

 disturbance in the general arrangement of the summit plates. 

 This might be expected, but it is certainly ver}^ remarkable that 

 the azygous proximal is divided also in species in which the 

 position of the anal opening is lateral or dorsal, and totally 



