HEMI ASTER GLOBULUS. 167 



so well shown on Plate 154 that no further description is necessary. The peri- 

 petalous fasciole is nearly or quite 3 mm. wide at the ends of the petals and 

 about 2 mm. wide at the middle of each interambulacrum, but on the bounda- 

 ries between ambulacra and interambulacra it is only 1 mm. in width. The 

 area it encloses is 23 X 17 mm. The posterior petals are nearly 6 mm. long 

 and almost half as wide; there are about a dozen pore-pairs on each side. Petals 

 II and IV are about 10 mm. long by 3 mm. wide and have 21-23 pore-pairs on 

 each side. Ambulacrum III is distinctly sunken within the fasciole; the peta- 

 loid portion is 13 mm. long by 2 mm. wide, with 21 or 22 pore-pairs on a side, 

 and does not reach the fasciole by about a millimeter. The abactinal system 

 is very compact and it is difficult to make out the sutures. The four genital 

 pores are very distinct but the ocular pores especially the posterior pair are hard 

 to make out. The madreporic genital (2) extends backward further than in 

 H . expergitus and distinctly separates the posterior genitals, 1 and 4, but oculars 



1 and V remain in contact. A step further and we should have such a system 

 as that shown in PI. 150, fig. 4, where the oculars I and V are separated but the 

 madreporic genital is not long enough to reach them. It would be but a slight 

 step further to a distinctly ethmolytic system such as occurs in most of the 

 family. 



The periproct is very small, nearly circular, 3.5 mm. in diameter, and situ- 

 ated high up on the posterior end of the test; its upper margin is scarcely 6 mm. 

 below the crest. All the actinal tube-feet are rather large but 5-6 at the lower, 

 posterior angle of the test are noticeably big. The sternum is about 14 mm. 

 long by 13 mm. wide posteriorly and is densely covered with primary spines, 

 all of which are broken. The peristome is about 6 mm. wide by 4 mm. long, 

 and its anterior margin is nearly 11 mm. from the front of the test. The plates 

 around the peristome (PL 150, fig. 5) are closely united to each other and have 

 also suffered injury so that it is difficult to be sure of their arrangement but appar- 

 ently the ambulacral plates 161 and 2 and Hal and 2, and IV61 and 2 and Val 

 and 2 are in contact so that interambulacra 1 and 4 are shut off from the peri- 

 stome. This is a striking difference from the condition in expergitus but I am 

 not sure that it is really the condition in globulus. 



The test is thickly covered with a fairly uniform coat of slender spines about 



2 mm. long, slightly spatulate at the end. On interambulacra 2 and 3 dorsally, 

 within the fasciole, on the sternum and around the mouth the primary spines 

 are much larger. Pedicellariae are relatively rare. Ophicephalous pedicel- 

 lariae are fairly common at the anterior end of the test but not elsewhere. The 

 stalks are 4-5 times as long as the heads. The valves (PL 146, fig. 11) are about 



