114 ART. 7. 1. IJIMA : HEXACTINELLTDA, IV. 



In the liemihaxactinose form (fig. 6) it is only one or at 

 most two of the exceedingly short principals that bear two widely 

 divergent terminals. A case of more than two terminals to a 

 principal has not lieen met with. 



The hexactinose form (fig. 2) is as common as, if not more 

 common than, the hemihexactinose. Uniterminal i"ays, whether 

 occurring in this or in the other form, are usually straight but 

 may sometimes be bent at base. One case that I observed of a 

 straight rayed hexactinose hexaster, in which two of the rays bore 

 each a small spine-like rudiment of an obsolescent terminal, 

 seems to be worthy of special note. 



In this species again it is a readily noticeable fact that the 

 largest of the oxyhexasters are found among those of the hexact- 

 inose variety. 



Exceptionally certain principals totally disappear together 

 with the terminals belonging to them. Thus arise peculiar de- 

 generate forms which under certain circumstances may have less 

 than six terminal points in all (figs. 4, o and 7). Fig. 4 is 

 evidently a case in which one entire axis is quite suppressed, 

 there remaining three uniterminal principals and a single biter- 

 minal one, all lying in one plane. In fig. o there are seen six 

 rays, but all these again lie in one plane ; they represent either 

 three biterminal principals or four principals of which two are 

 bi terminal and the rest uniterminal. Fig. 7 is an extreme case 

 of the reduction ; there are left only three rays, assumably repre- 

 senting as many half-axes each composed of a principal continued 

 into a single terminal. 



The most characteristic spicule of the species is the macro- 

 discohexaster (PI. VIII., fig. 27 ; PI. IX., fig. 8). This is strik- 



