110 CATALOGUE OF THE BLASTOIDEA. 



remain separate, or, in other words, where we have ten distinctly visible openings." 

 This passage is illustrated by a diagram of Pentremites melo, in which there are 

 eleven openings shown just as in that of Pentremites sulcatusl As a matter of fact, 

 however, there should only be nine, for the anus is confluent with the two posterior 

 spiracles, as seen in our fig. 14 on PI. VII. But this is not the only error in 

 Hambach's diagram. The spiracles are placed too far from the peristome, and are 

 consequently represented as further apart than they really are. His statement that 

 the spiracles have to remain separate because the zigzag plicated integument is too 

 narrow to surround them fully is another instance of his teleological arguments. 

 But they are not worth discussing, because this supposed elastic integument is 

 nothing else but the crenulation of the food-groove on the median line of the ambu- 

 lacrum and of its lateral branches between the side plates, as pointed out on p. 59. 



Cryptohlastus has an hydrospire-plate and a compound anal spiracle; Acentrotremites 

 probably has an hydrospire-plate, but the anus is distinct from the posterior spiracles. 

 Neither Schizoblastus, Elceacrinus, nor any of the Troostoblastidee have any hydro- 

 spire-plate. But the Irish species of Schizoblastus (PL VIII. fig. 9 ; PI. XVI. fig. 12), 

 and also Troostocrinus Beinwardti and T. Grosvenori, have an anal spiracle, though iu 

 Elceacrinus, in the American species of Schizoblastus, and in the genera Triccelo- 

 crinus and' Metablastus, the anal opening is distinct (PI. III. figs. 1-3, 11, 15; 

 PI. VI. fig. 16; PL XVIII. figs. 15, 16 ; PI. XIX. figs. 15, 16). In all the four 

 types last mentioned the spiracles are formed upon essentially the same plan. The 

 ambulacra are narrow and often diminish in width as they approach the peristome, 

 as is especially the case in Elceacrinus (PL XVIII. fig. 16) ; while the proximal end 

 of the radial sinus encroaches on the sides of the deltoids so as to form a more 

 or less conspicuous opening. One edge of this opening is bounded by the narrow 

 ambulacrum, which soon comes into contact with the deltoid so as to close the 

 distal border of the spiracle. But as the side plates only rest upon the Jancet- 

 plate without projecting beyond it, they take no part in the formation of the 

 spiracle, and do not directly cover in the hydrospire-canal as is the case in the 

 Pentremitidae. They have nearly the same relation to the spiracle as in Crypto- 

 hlastus and Acentrotremites (PL VII. figs. 14, 15 ; PI. XIII. fig. 19), except that the 

 lancet-plate between them is not incised to form the edge of the spiracle as is 

 the case in these two genera. 



It may also be noted that the spiracles of such a form as Metablastus lineatus 

 (PI. III. figs. 14, 15) are very closely similar to those of some species of Oropho- 

 crinus, e. g. of 0. verus (PL XV. figs. 2, 3). In each case the spiracle is the 

 proximal end of the radial sinus, which is closed distally by the ambulacrum meeting 

 its walls. But in the former type the hydrospire-canal, which is continued out- 

 wards beneath the ambulacrum, communicates with the exterior through the hydro- 

 jpire-pores, which are really only constricted portions of the sinus not closed by 



