1886.] NATURAL SCIENCES OF PHILADELPHIA. 113 



manner. In these genera it is exceedingly difficult to determine 

 whether the branches are armlets or pinnules, a question which 

 cannot be decided definitely until we know where the genital 

 glands are located. It is probable that in the genus Cyathocrinus, 

 at least in its Carboniferous form, all branches were artos. We 

 found in G. multihrachiatus^ along the three or four proximal 

 arm joints, outside the adambulacral or side-pieces, small plates 

 in rows of from four to six pieces (PI. 4, fig. 7 a, 6), succeeding 

 each other longitudinally, which perhaps took the place of the 

 genital pinnules. No such plates, however, were represented in 

 Baryc?'inits^ in which the ventral grooves are comparatively 

 narrow. Cyathocrinus longimanus has no regular pinnules, but 

 certain sabre-shaped appendages, composed of five segments, 

 which from each side infold over the ventral furrow, covering it 

 completely. These appendages which we (Rev., i, p. 25) erro- 

 neously took to be rudimentary pinnules, ^ perhaps correspond 

 with the so-called " pinnules " of the Cupressocrinidte and Blas- 

 toidea. 



Another good distinction between the Cyathocrinidae and 

 Poteriocrinidae is offered by their mode of articulation. The 

 radials of the former have horseshoe-like facets for the brachials ; 

 in the Poteriocrinidae they are more or less truncate along the 

 upper margin, and united with the brachials by a transverse ridge, 

 occupying a median line and frequently their entire width. The 

 middle part of this ridge is pierced b^^ an axial canal, and there 

 is a kind of muscle plate with more or less conspicuous fossae. 

 The outer edge of the plate is dentated and evidently was occu- 

 pied by ligamentous bundles. In the Cyathocrinidae the arms 

 are always bifurcating, and their branches are given oflf at close 

 intervals ; those of the Poteriocrinidae are frequent!}' simple from 

 the brachial bifurcation upwards, but when bifurcating, the 

 divisions are o;iveu off irregularlv, and branches and main arms 

 bear pinnules alternately from every joint ; there being no S3'zy- 

 gies. The arm joints of the former, with the exception of Bary- 

 crinus, are composed of long, slender joints with almost parallel 

 sutures ; those of the latter are shorter, heavier and strongly 

 wedge-shaped, even interlocking. The same mode of articulation 



^ A similar interpretation at that time was given by us of the anibulacral 

 plates covering the arm furrows of Cyathocrinus iowensis, a mistake which 

 w as rectified by Carpenter (Chall. Rep., pp. 63-GG). 



