378 PROCEEDINGS OF THE ACADEMY OF [1890. 



cally corresponding to an ordinary brachial," which gradually in 

 its palaeontological development passed down from above the radials 

 to the basals, and between the radials. To the former plate he gives 

 the name " radianal,^' which we think an excellent one, and we ac- 

 cept it as being more appropriate than azygons plate. To the latter 

 he applies the name " brachianal," a term which becomes meaning- 

 less if the plate proves to be an interradial. For this plate we re. 

 tain the name analplate, as we hold it to be the true homologue of 

 the anal plate in the Antedon larva, and the homologue oitl^e first 

 anal plate of Actinocrinus. 



In our diagrams we have marked this plate X, and we find it 

 convenient to refer to it as such in this paper. We apply the letter 

 R to the radials, and when these are compound, we distinguish the 

 lower, non-arm-bearing section as R'. 



To understand the relations of the anal plates, we must at first 

 direct attention to the structure of the radials of the earlier Fisiulata, 

 which in all Loiver Silurian genera, probably with the exception of 

 Ottaivacrinus, ^ differ essentially from those of the Upper Silurian 

 forms. It is very curious that in the former the radials in from one 

 to three of their rays are compound, i. e., constructed of two segments 

 or parts, which are closely united by a horizontal suture, and in the 

 organization of the Crinoid count as one plate. Compound radials 

 are found in the Heterocrinidae, Anomalocriuidae, Hybocrinidae, and 

 in the earliest Cyathocrinidae. In some of these families they are 

 restricted to the right posterior ray, in others one or two of the 

 other rays also may have compound radials. But we know of no 

 case in which the compound structure extends to all five rays, at 

 least two of the plates being always simple. Among the Hetero- 

 crinidae, as a rule, the right posterior and the right and left antero- 

 lateral radials are compound — exceptionally the anterior one in 

 ])lace of the left antero-lateral — while Anomalocrinus and Ohiocruius 

 have but two, and loerinus, Meroerinus, Hyhoerinus and Hoplocriniis 

 only one. It is further worthy of note that when there are several 

 compound radials, the corresponding parts are of nearly equal size; 

 while the segments vary considerably among themselves in the 



1 Carabocrimis, which also we had referred to tlie Fistulata, is pos-ibly a 

 Cystid. We were recently informed by Mr. Walter R. Billings that it may have 

 hydrospires. He was kind enough to send us carefully prepared diagrams of this 

 genus and of Hybocrimis,'^Q-^\r\Q^ the structure of the ventral disk, for which he 

 has our hearty thanks. 



