JAPANESE ASTEROIDEA. 303 



described in this paper ; in my specimen it is 2-2.5 mm. long. 

 The second adambulacral spine is mostly about one-third longer 

 than the first, and is also flattened ; it is, however, nearly 

 straight and presents only a shghtly curved border toward the 

 ambulacral furrow. The third spine is perfectly straight and 

 nearly round, and is as much longer than the second spine as 

 •• the latter is than the first. On many of the adambulacral plates 

 there are one or two forcipiform pedicellarisß of varying size 

 close to the third adambulacral spine, generally on its adcentral, 

 but sometimes also on its outer side (PI. VI, fig. 98, 100). There 

 are in addition a few more very small spines on the adambula- 

 cral plate, mostly on its outer part, but also more or less on the 

 adcentral side near the pedicellarise. The first four or five 

 adambulacrals at the mouth end mostly bear four instead of 

 three spines ; but in other respects they are exactly similar to 

 the rest. 



Mouth-plates. — The mouth-plates are comparatively small. 

 Each plate bears on its actinal ridge a series of five or six 

 spines, of which the one at the mouth end is stout and long, 

 and with its fellow forms a conspicuous oral armature projecting 

 towards the mouth. The other spines are of similar shape but 

 become smaller towards the abcentral end of the plate (PI. VI, 

 fig. 97). On the abactinal side of the large oral spine and parallel 

 to the ambulacral furrow is a series of three forcipiform pedicel - 

 lariœ, of which, however, the middle one is frequently transformed 

 into a flattened spine with truncated end. The pedicellaria 

 nearest the mouth is usually about half as long as the oral 

 spine ; the second one is smaller and the third is smallest. 

 When the middle pedicellaria is transformed into a spine, this is 



