242 CATALOGUE OE THE BLASTOIDEA. 



spiracle in G. Norwood! is bordered on the outer side by a solid projection formed by 

 a part of the deltoid plate 1 . Messrs. Meek and Worthen suggested that the anal 

 plate consisted of three pieces 2 , but we have not met with any evidence confirming 

 this view ; and we think that the characters of the internal casts to which they refer in 

 support of it are simply due to the presence of the two hydrospire-canals at the sides 

 of the rectum, as shown on PL VII. fig. 7. 



The ambulacra of Granatocrinus are always long, curved, narrow, and reach to the 

 base, the calyx usually resting, when placed in its natural position, on the distal ends 

 of the ambulacra (PL II. figs. 32-35; PL IX. figs. 1,2, 11, 12). This appears to be 

 an essential character of the genus, and is one of the points in which it agrees with 

 Schizoblastus (PL VI. fig. 15 ; PL VIII. fig. 10 ; PL XVI. fig. 13), and also with all the 

 members of Roemer's group Elliptici. The side plates in the ambulacrum of Grana- 

 tocrinus, unlike those of JPcntremites, do not lie against the sides or edges of the 

 lancet-plate, but actually rest upon it, so as to conceal more or less of its surface 

 (PL VIII. figs. 15, 21 ; PL IX. fig. 16 ; PL X. figs. 11, 12). They vary in number 

 according to the species (from twenty to eighty), and are usually transversely elongated. 

 The ambulacra are at times deeply impressed within the prominent edges of the radial 

 sinuses. In G. Nonvoodi, and in all the British species with one exception, the lancet- 

 plates almost entirely fill up the radial sinuses (PL X. figs. 8-12). "We have come 

 to the conclusion that in Granatocrinus, as restricted by us, there is no under lancet- 

 plate, as there is in Penfrcmites and Orophocrinus, and we are glad to find ourselves 

 in accord with Messrs. Wachsmuth and Springer on this point 3 . There is a good deal 

 of variation in the formation of the ambulacral pores. In G. orbicularis they are 

 enclosed between the margin of the radial plates, the lower margin of the side plates 

 above, and the upper oblique margin of the outer side plates, and are, roughly 

 speaking, triangular in shape. The result of this is that the pores are excavated 

 wholly in the outer side plates (PL IX. fig. 16). In G. cllipticus a slightly different 

 arrangement is met with. Here the pores are excavated out of the side plates 

 themselves, the upper edge of the outer side plate being quite straight, and not at all 

 incised (PL VIII. fig. 21). In G. campanulatus the pores are scarcely excavated in 

 the ambulacral plates at all, but are almost wholly formed in the edges of the radial 

 plates (PL VIII. fig. 15). There also appear to be slight modifications in the arrange- 

 ment of the sockets placed on the side plates. In G. orbicularis they terminate 

 narrow grooves which arise from the lower sutures between the side plates and the 

 outer side plates, running almost straight on to the centre of each plate (PL IX. fig. 16). 

 In G. e/lijjticus, on the other hand, the grooves communicating with the sockets arise 



1 Boston Journ. Nat. Ilisl. L863, vol. vii. no. 4, p. is.'i. 



- Ilrport Geol. Survey Illinois, lb~'-i, vol. v. p. 465. 



3 Revision of the Fakuocrinoidca, Part II. 1881, (. L9. f. <l. 



