102 CATALOGUE OF THE BLASTOIDEA. 



with denticulate edges. This is the cast of the hydrospire-canal with its roof of side 

 plates and intervening pores. It is sometimes hollow, and its cavity communicates 

 with those of the hydrospire-slits. The casts of the two hydrospire-canals at either 

 side of each interradial approach one another at the summit, and their proximal ends 

 represent the spiracles. In the anal interradius they are seen to he quite distinct 

 in origin from the cast of the anal opening (PI. III. fig. 12), though they are scarcely 

 distinguishable from it at the summit (fig. 10). The same difference appears on 

 comparison of the two figures 15 and 17 on PL XII. In the first case the summit 

 has been ground away far enough to expose the hydrospire-folds in the spiracles, 

 except in those at the sides of the anus, which are well separated from it. But in 

 the other specimen, which has been less ground, the two septa in the anal spiracle 

 are almost entirely removed. 



Hambach 1 has recently attempted to describe the formation of the spiracles in 

 the typical 1'cntremitcs, with which he associates two species that we refer to the 

 Troostoblastidse, viz. : — P. Wortheni and P. Beinwardti. But we cannot think he has 

 been very successful, as some parts of his description apply to but a few species of the 

 typical Pentremites only. He says for example, " The first division would comprise all 

 those species in which the horizontal portion of the deltoid piece is very narrow, the 

 sinus to both sides in the deltoids and lancet pieces comparatively large, and so 

 surrounded by the zigzag plated integument that two of the so-formed openings 

 appear externally only as one (see fig. 5a)." In this figure the spiracles are repre- 

 sented as formed by the apposition of notches in the lancet-plate and in the deltoids 

 respectively, very much in fact as is shown in our figure of Cryptoblastus melo 

 (PL VII. fig. 15) ; but there is no indication that the side plates take any part in 

 the formation of the spiracles. In fact Hambach states in an earlier paper a that 

 the lateral concave furrow of the lancet-plate " forms, in connection with the corre- 

 sponding portion of the deltoid piece, the ovarian aperture." It would appear from 

 his figure that the side plates take as little part in the formation of the spiracles of 

 Pentremites as they do in Cryptoblastus melo (PL VII. fig. 15). We cannot pretend to 

 make a general statement about all the species of Pentremites found in America, but 

 so far as what Hambach calls the typical species are concerned, viz. P.Jiorcalis (i. e. 

 P. Godoni), P. sulcaius, and P. piriformis, his description and figures certainly seem 

 to be somewhat incorrect ; though we believe that they are more applicable to Troosto- 

 crinus Beinwardti and to Hall's Pentremites Wortheni. In all the typical Pentremites 

 the side plates form a part of the distal boundary of the spiracle as shown in P. 

 sulcatus (PL I. fig. 10) and also in Hambach's own species P. hemisphericus (PL XVI. 

 fig. 21). The same thing appears in the dissected ambulacrum of'y j . piriformis shown 

 on PL I. fig. 6. The hollow in the surface of the deltoids on either side of the ambu- 



1 Trans. St. Louis Acad, Sci. L884, vol. iv. no. 3, p. 544. 

 -' Ibid. L880, vol. iv. no. I, p. 1 19. 



