124 PROF. BELL ON NEW SPECIES OF ASTEROIDEA. [Jan. 17, 



actinic peculiarity ; and the decision can only be given with the aid of 

 evidence which proves or disproves the idea that the species can and 

 does undergo division. Such evidence is here afforded by the spe- 

 cimen in the national collection ; for it has the arms of different 

 lengths ; this, of course, points to some of the arms being younger 

 than the others, or, in other words, as having arisen by gemmation 

 after division. 



The heteractinic, rather than the sexradiate, condition should 

 therefore be regarded as a part of the diagnosis of the species. 



Fromia tumida, n. sp. (Plate VI. figs. 4, 4 a.) 



Arms and disk more swollen than in most of the species of the 

 genus. R = 4.5, r= 14; R = 35, r = 11-5, orR= 3'2r (about): 

 breadth of arms at base 14'5, 12*5 millim. All the plates on the 

 actinal surface, with the exception of those of the marginal series, are 

 closely covered with stout spines, from which the spines of the adara- 

 bulacral series are distinguished by their smaller size. The whole of 

 the abactiual surface is covered by squarish-headed granules, which are 

 all of very much the same size, and are all very regularly distributed ; 

 they are arranged in elliptical aggregations, two or three rows of 

 which extend along the back of the ray, or they are placed between 

 these aggregations ; in the former case they are somewhat more 

 closely packed. The intermediate spaces, in addition to the 

 granules, present a certain number of pores, which, however, are, 

 in comparison with other species of this genus, rare. The madre- 

 poric plate is rather nearer the margin than the centre of the disk, 

 and is very similar to the same body in F. milleporella. There are 

 about seventeen marginal plates in both the actinal and abactinal 

 series ; and both sets are covered with granules of a fiiir size ; and the 

 marginal granules of each set are to be easily distinguished from the 

 more central ones which they surround. For some way along the 

 surface of the arm the dorso-marginal plates are deeper than long ; 

 the infero-marginal plates are much more nearly square. There 

 seems to be a large pore at the proximal angle between the two sets 

 of marginal plates ; but as the two specimens on which this descrip- 

 tion is drawn up have both been dried, it is not possible to speak 

 definitely on the point. The same remark will apply to the pore at 

 the proximal angle of the actinal edge of the infero-marginal plates. 



The adambulacral spines appear to be arranged in three longi- 

 tudinal rows. They are subequal in size, and diminish very slightly 

 as they approach the free end of the xny. The space between these 

 spines and the marginal plates is completely occupied with short 

 stout spines, set in tufts on small plates. No pedicellarise. 



Ceylon. Presented by M. Kelaart. 



EXPLANATION OF PLATE VI. 

 Fig. 1. Scales of Calvasterias antipodum, X 4 : 1. 



2, Abactinal surface of Cribrella minuta, X 2 : 1. 



3, 3«. Actinal and abactmal surfaces of Mithrodia victorice, X 2 : 1. 



4, 4a. Actinal and abactinal surfaces of Fromia tumida, showing the 



arrangement of the plates and spines, X 3 : 2. 

 46. Portion of actinal surface, more highly magnified. 



