80 THE CRINOIDEA CAMERATA OF NORTH AMERICA. 



Compsocrimis, — both from the Lower Silurian, — have uniserial arms; Peri- 

 gl//ptocrinus, and probably all later Melocrinid^, biserial. Among dicyclic 

 Camerata, we find well defined biserial arms already in the Trenton and 

 Hudson River groups, along with uniserial, the former being perhaps in the 

 majority ; but the latter are continued to the lower part of the Devonian by 

 two species,* and these, together with the species of DicJiocrimis, are, so far 

 as we know, the only representatives of the Camerata with a single row of 

 arm plates that survived the close of the Silurian. 



In the Fistulata, the biserial arm structure was introduced just before the 

 close of the Carboniferous, but only in a limited way. True biserial arms 

 only occur in Graphiocriniis,-\ Eupachycrinus, Cromyocrinus, Hi/dreionocrinus, 

 Erisocriniis^ Stemmatocrinus, and Encrinus, but the majority of species have 

 either quadrangular or cuneate arm plates, and in some of them only the 

 tips of the arms begin to interlock. In this group the biserial stage at no 

 time became a constant character, not even in the Triassic. Encrinus lilii- 

 fonnis has perfectly biserial arms, while the arms of Encrinus gracilis are 

 uniserial, and composed of quadrangular plates. 



The arms of the Articulata, not only in Palaeozoic pinnuleless forms, 

 but also in the Neozoic pinnule-bearing ones, are uniserial without exception. 



The pinnules, as happily expressed by Carpenter, are repetitions of the 

 arms on a small scale ; and in their organization morphologically, and to 

 a large extent physiologically, closely resemble ordinary arm branches. 

 They are short branchlets given off along the sides of the arms, but rarely 

 reaching their tips, and are usually more slender, and composed of longer 

 joints. The pinnules differ from arms in containing the fertile portions of 

 the genital glands, while the arms lodge the genital cords. Like the arms, 

 they have ambulacral furrows fringed with cilia, by means of which particles 

 of food coming in contact with them are carried along the grooves to the 



* Oelilert's new genus Biamenocrinus, and " Rhodocrinus " gonatodes Miiller (both from the lower 

 Devonian), which perhaps belong to one genus, have dichotomizing arms, composed of short, quadrangular 

 plates. 



f It is doubtful if the name Qraphiocnmis can be retained, according to the general rules adopted by 

 naturalists, as it was incorrectly defined by de Koninck and Le Hon. The type has small infrabasals hidden 

 by the column, and the position of the anal plate is materially different from that given by the Belgian 

 writers, which probably was not known to Trautschold when he proposed the genus Phialocrimis. The anal 

 plate of Graphiocrinus encrinoides, de Koninck's type, rests directly upon the truncated posterior basal, as 

 we observed in a fine specimen in our collection, exactly as in Pliialocrinus. There is, however, a slight ob- 

 jection to Trautschold's name, which was preoccupied by Eichwald (Lethsea Rossica I, p. 578), but the 

 genus was founded merely upon fragments of column. That Pliialocrinus patens has two costals, and Graphio- 

 crinus encrinoides but one, may not be of generic importance, as the two plates of the former are equal to the 

 one of the latter, which evidently form a syzygy. 



