HYALASCUS SAGAMIENSIS. 95 



sometimes a more pronounced and somewhat angular bending at 

 base. The latter condition is one which would arise directly 

 from the biterminal state by total atrophy of one of the termi- 

 nals, w^hile the former condition represents transitional phases of 

 a uniterminal ray towards becoming perfectly straight at base. 



Noteworthy seem the degenerate oxyhexasters with less than 

 six terminal points, such cases being certainly quite rare. In 

 PI. VII., fig. 8, I have shown a case which in view of the shape 

 might be called an oxystauraster. There can be no doubt what- 

 ever that this spicule was derived from a hexactinose oxyhexaster 

 by comj)lete suppression of two opposite rays occupying the posi- 

 tion of an axis. 



The discohexasler (PI. VIL, fig. 5) is moderately common 

 near the gastral surface. It is probably not totally wanting in 

 the parenchyma generally. It is rather small in size, spherical 

 in shape and of an exceedingly delicate nature. Diameter, 80- 

 90 n. A spherical central node is sometimes distinctly percep- 

 tible and sometimes not. The six principals are short, being only 

 about 3 It. long ; their outer ends seem to be simply truncate, 

 instead of forming a disc-like expansion. The terminals, of which 

 there are usually only three and exceptionally four to each 

 principal, are very fine filaments which thicken somewhat towards 

 the outer end. The small number of the terminals radiating in 

 all directions seems to be characteristic, forming the most impor- 

 tant diagnostic feature by which the present species can be dis- 

 tinguished from Hijalascus similis. The terminal discs are quite 

 small ; in lateral view they appear arched like a watch-glass. 

 Their marginal dentation could not be brought into view. The 

 terminals break off easily at a certain distance from the base, so 



