MORPHOLOGICAL PART. 109 



crinidae, the covering pieces are generally exposed in the calyx, while in the 

 Actinocrinidas they are as a rule hidden from view. But occasionally the 

 opposite is the case, and even among species of the same genus. Adinoaimis 

 stellaris, from Belgium, has well defined covering pieces passing out from the 

 outer edges of the orals ; while almost every other species of Actinocrinus has in 

 place of them radial dome plates of the first, second, and third order, accord- 

 ing to the number of bifurcations in the calyx. Phi/setocrinus, which is but a 

 modified Acimocrmus, has in some species orals and radial dome plates ; w^hile 

 in others the whole ventral surface is covered with small, irregular pieces. 

 In Actinocrinus miiUiradiatus^ the entire tegmen consists of only a few unusu- 

 ally large interradial plates, which interlock with those of the dorsal cup. 

 But the interradials in most of the Actinocrinidse pass insensibly into the 

 tegmen, there being no dividing line ; while in Batocrinus generally, but 

 not always, the interradials of the dorsal side are distinctly separated from 

 those of the ventral side by the arching brachials, — a structure which led us 

 at first to suppose that the plates of the two hemispheres were morphologi- 

 cally distinct. 



Similar variations occur among the plates of the tegmen in the PI a ty- 

 crinidaB and Hexacrinidae. In some (3f their species the pavement is made up 

 entirely of massive plates, in others of comparatively thin pieces; while in 

 still others the ventral surface is occupied almost exclusively by the orals. 

 In both groups it is absolutely impossible to draw a dividing line between 

 interbrachials and interambulacrals. The plates constituting the first row, 

 which generally consist of three pieces, are peripheral and partly interam- 

 bulacral, and those of the succeeding rows strictly ventral. The plates of the 

 second and higher rows, when such are present, interlock wdth those of the 

 first row, like the interradial plates of the dorsal cup in an Actinocrinus. 



The condition of the ventral pavement in the Melocrinidae, Rhodocrinidae, 

 and Thysanocrinidae is similar to that in the Actinocrinidae. Their lower 

 interbrachials are definitely arranged, and there is no line of demarkation 

 between the plates of the two hemispheres, except that produced by the 

 arms, which pass out between them. In the Reteocrinidae, as in most of the 

 other Silurian Camerata, the wdiole ventral surface is covered by minute, 

 irregular pieces, and similar plates, with a few somewhat larger ones scattered 

 among them, are interposed between the rays from the basals up. 



It seems to us perfectly clear from the structure that all interradial and 

 interaxillary plates, not only in the Camerata, but also in all recent and fossil 



