364 PROCEEDINGS OF THE ACADEMY OF [1890. 



Ever since we discovered that tlie ventral surface of Taxocr'uius 

 is a true disk, we became convinced that the views heretofore held 

 by us respecting vault and disk, must be greatly modified or alto- 

 gether abandoned. We have since given considerable thought to 

 the subject, and in the latter part of last year laid the results before 

 Prof. Alex. Agassiz and Dr. Carpenter, who both encouraged us to 

 continue our researches in that direction. Prof. Agassiz informed 

 us that he had come to quite similar conclusions respecting those 

 plates from his own studies, and Dr. Carpenter with his usual liber- 

 ality gave us valuable hints and explanations, and now agrees with 

 us on this question in all essential points. 



The ventral pavement of the Camerata is composed of interradial 

 plates which, as before explained, form a continuation of the plates 

 from the dorsal cup, and these meet with the orals, where they are 

 represented. Frequently also the covering plates of the ambula- 

 cra take part in the pavement, and a set of plates to which we gave 

 the name " radial dome plates." As the latter plates were supposed 

 to form an integral part of the vault, overlying the ambulacra, they 

 deserve especial attention. They were regarded by us, and also by 

 Carpenter, as the actinal representatives of the radials, but later in- 

 vestigations have shown that they are highly differentiated cover- 

 ing pieces. The plates appear in two forms. They are either fol- 

 lowed by series of covering pieces, and pass out from between two 

 orals, as in most species of Flatycrinus ; or they are in a certain 

 sense isolated plates, surrounded by other "vault" p»lates, and suc- 

 ceeded by similar plates of higher rank, as in the case of the Actin- 

 ocrinidae with subtegminal ambulacra. The former case is quite 

 readily perceived if we examine the Silurian forms, in which, when 

 their ambulacra are exposed, two regular series of alternating plates 

 pass outward directly from the orals. Here either the radial dome 

 plates are wanting, or they must be represented by the proximal 

 covering plates. In the later Platycrinidae, the covering plates 

 throughout lose much of their original character, being, as a rule, 

 more extravagantly developed than in the earlier forms, and the 

 proximal plates of each ambulacrum are larger than the succeeding 

 ones, so as to obscure the alternate arrangement. 



The " radial dome plates " of the Actinocrinidae and allied forms 

 are generally larger than any of the surrounding plates, often nodose, 

 and sometimes extended into long spines. They are not followed 

 by covering pieces like those of the Platycrinidae, and, unlike them, 



