374 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



the 2 Philippine examples. Doctor Koehler does not compare major 

 with japonica in the references given above. 



The example from station 5251 is remarkable for having numerous 

 actinal pedicellariae (resembling split granules with from 2 to 4 

 jaws) from 3 to 6 times the diameter of the actinal granules. These 

 are largest and most numerous in the actinal interradial areas and ex- 

 tend about half the length of the ray, being found on the actinal inter- 

 mediate plates, with a very few also on the proximal inf eromarginals. 

 There are 2, 3, or 4 furrow spines in this specimen and 2 subambul- 

 acral spines (often 3 proximally) . When there are 3 the middle spine 

 is the longest, and when there are 2 usually the aboral is the longer. 

 For the distance of about 6 inferomarginals from the end of ray only 

 1 subambulacral reaches any size. The outer edge of the plate is 

 occupied by 1 (or occasionally 2) series of 4 to 6 granules, or granules 

 and pedicellariae, the former being uniform in size with the actinal 

 granulation (pi. 92, fig. 2), 



In the specimen from station 5159 there are 3 furrow spines at the 

 base of the ray, then either 2 or 3, and finally only 2 on the outer 

 half or third of the ray. There are usually 2 subambulacral spines, 

 the adoral a trifle the larger, but near the end of the ray frequently 

 only 1 reaches any size. External to the subambulacral spines, which 

 are flattened, slightly tapering, round-tipped, and a trifle shorter and 

 stouter than the furrow spines, are 2, 3, or even 4 longitudinal series 

 of granules belonging to the adambulacral plates. In both specimens 

 the mouth spines are longer than the corresponding adambulacral 

 spines, and instead of being blunt are sharp. The granules surround- 

 ing the papular pores are only a trifle larger than the others. 



Fromia monilis Perrier (1875, p. 179) has never been figured, and 

 Doctor Koehler did not see the type when working up his Indian 

 material (1910a, p. 140). No one has compared it. therefore, directly 

 with F. Tfrmjor or F. japonica. It apparently has alternating large 

 and small superomarginals and a narrow abactinal area. In fact, 

 the description fits japonica fairly well up to the adambulacral plates, 

 which have 2 furrow spines and regularly a single subambulacral, 

 which is described as being much larger than the furrow spines, 

 cylindrical, and terminated by an obtuse point. These spines, isolated 

 on each plate, form the second row. On the distal half of the ray 

 of the type of japonica there is only one subambulacral spine, but it 

 is shorter and thicker than the furrow spines. 



Fromia japonica differs from F. milleporella in having longer and 

 slenderer rays, regularly alternating large and small superomarginal 

 plates (at least on outer part of ray), and slightly larger granules 

 surrounding the papular pores. 



Type. — No. 743, Museum of Comparative Zoology. 



Type-locality. — Japan. 



