JAPANESE ASTEEOIDEA. 331 



tion. It can easily bo seen that the radiating elevations along 

 the arms are due to the presence of the digestive system which 

 is regularly petaloid in shape. 



The pedicellaria) are present both on the actinal and 

 abactinal sides as well as on the marginal plates. Those on the 

 actinal sides are very few in number, there being mostly only 

 one in each interradius and that one is not always present. It is 

 closely similar in form and structure to the pedicellaria? found in 

 similar positions in Mediader brachiafus, but so far as I have 

 observed, it is tricuspid and is mostly situated on one of the 

 ventrolateral plates near the mouth (Pi. XI, fig. 174). 



The pedicellarife of the abactinal side are all val va te, being 

 much more elongated transversely than high. They are borne on 

 the abactinal plates, and are mostly, though not invariably, formed 

 by the homologues of the peripheral granules (PL XI, figs. 175, 176). 

 In the specimen with 1\:=50 mm. above mentioned I counted 

 as many as 40 pedicellariae on the whole abactinal surface. 



Several of the marginal plates, both inferior and superior, 

 bear pedicellaria3 which are closely similar in form and structure 

 to those of the abactinal surface, but they are not very constant 

 in position and number. 



The pedicellariae of this species appear to be subject to much 

 variation in number, so that Sladex's statement on this point is 

 perfectly intelligible. 



As to the adambulacral armature, the spines of the furrow 

 series vary between five and seven, and those of the second series 

 are four or five in number, very rarely six (PL XI, fig. 174). 



The mouth-plates are also subject to variation in size and 

 general appearance, being tolerably large and conspicuous in the 



