440 = Mr. F. J. Bell on a new Species of Archaster. 
It differs from <A. rigida in the small proportions of the 
calices, from A. varia in the non-cellularity of the corallum 
and the shallowness of the calices, from P. Hspert in the 
small size of the calices and of their parts and in the. slight 
denticulation of the mural ridge by the septa. 
The only coral alluded to in the paper on Ascension in the 
Phil. Trans. for 1699 could hardly be this species, though 
some Astreid is possibly indicated by the very vague and 
unscientific description there given. 
XLV.—Description of a new Species of the Genus Archaster 
from St. Helena. By ¥. Jerrrey Beuu, M.A. 
In connexion with the foregoing notes on the fauna of the 
island of Ascension it seems to be of interest to present an 
account of a magnificent species from the island of St. Helena, 
two specimens of which were, several years ago, presented to 
the Trustees of the British Museum by Mr. J. C. Melliss, but 
to which the donor makes no reference in his valuable work . 
on that island *, and which have not as yet been described. 
Archaster magnificus, 1. sp. 
Rays five. R=207,138; r=50,37. Breadth of arms at 
base 57, 36; at middle 36, 20; breadth of paxillar area at 
middle 17, 10°5. 
Disk rather flat, the median portion of the arms elevated, 
and the paxille along the middle line arranged in longitudinal, 
and not, as at the sides, transverse rows. Anus obscure. 
The tips of the paxille consist of a tuft of about fifteen 
cylindrical spinules, closely appressed together, and not forming 
any kind of fringing crown. ‘There are, in one specimen, about 
seventy-four, and in the other about sixty-four, supero- 
marginal plates ; these are covered with granular scales, which 
may become elongated, or converted at the lower edge into 
flattened spatulate spines, which do not appear to stand erect, 
but to le along the side of the arm. The plates themselves 
are placed altogether at the margin of the arm, and are some- 
what higher than broad. Near the tip of the arm the spines 
on these plates may disappear, while the plates themselves 
take a more dorsal position. The infero-marginal plates 
are thickly covered with stout flattened spines, which are 
much better developed than in A. angulatus. The spines are . 
frequently cut off square at their tips, and are so closely 
packed as to destroy any distinct appearance of regularity of 
distribution ; here and there, however, it is possible to detect 
* J.C. Melliss, ‘St. Helena’ (London, 1875). 
