332 s. GOTO : 



smaller two of my specimens, and less so in the largest. 

 A closer examination, however, reveals that this is due more 

 to the cliaracter of their spines than to their actual size, the 

 former being closely similar to the armature of the ventro- 

 lateral plates in the largest of my specimens, while in the other 

 two the spines are larger and more pointed. There are some ten 

 or more flattened, robust spines on the furrow margin of each plate, 

 four or five short, prismatic ones on the border facing the first 

 adambulacral plate, and some seven spines along the interradial 

 border. Of the last mentioned series the last one or two spines 

 next tlie mouth are especially large and pointed (PI. XI, fig. 173). 



With regard to the abactinal plates, they are, as remarked 

 by Sladex, strikingly like paxilloo. The coronal granules are 

 somewhat elongated ana the centrals, when present, are mostly 

 larger than the peripherals. On one of the larger plates there may 

 be as many as 15-20 peripherals and 10 centrals (PL XI, fig. 175, 

 170). The latter may, however, be totally absent. The bases of 

 the abactinal plates are connected by elongated internal ossicles, 

 between which lie the papular pores, exactly as in Mediaster. 

 The number of these ossicles radiating from a single plate is 

 mostly six. 



Specimens in S.C. 



JPentfff/onaster inisakiensis, u. sp. 

 (PL Xni, figs. 194-202.) 



The body is regularly stellate, the arms are tolerably long, 

 bluntly pointed and in my specimen strongly curved upwards in 

 the distal half (PL XIII, fig. 194, 195). This last point must 

 however depend much on the condition in which the animal was 



