REPORT ON THE REEF-CORALS. ing 
distinct on the marginal parts, close, thin, and regularly arranged in parallel lines; but 
which, as the growth proceeds, become gradually thicker and closer, and eventually fuse 
to form a broad, rounded, solid wall. The costal area is also peculiar owing to the sub- 
concentric sharp edges which mark it, and which would appear to have been produced by 
separate periods of growth. 
Through this form the genus Agaricia is closely related to Siderastrea. 
Locality.—Levuka. 
Family CycLosErID&. 
Genus 1. Cycloseris, Milne-Kdwards and Haime. 
Cycloseris et Diaseris, Milne-Edwards and Haime, Cor., iii. pp. 49, 54. 
a Ke Dunean, Rev. Madrep., pp. 149, 150. 
I have placed Diaseris as a synonym of Cycloseris, since I am convinced that the 
curious forms which have been referred to that genus are nothing more than broken 
pieces of specimens of Cycloseris, which in this broken condition have continued living, 
and have, in part or wholly, repaired the injury. 
The early view which was held by Milne-Edwards and Haime as to the constitution 
and formation of these curious forms, namely, that in the young state the specimens have 
the form of separate pieces or lobes, and that these unite irregularly durmg growth, has 
long ago been proved by Semper to be untenable ; and indeed one is inclined to wonder 
that such an opinion, with all that it implies for the formation of those structures, should 
have been advanced by such writers as Milne-Edwards and Haime. 
Tenison-Woods has already noticed the extremely close relationship between the two 
genera, though the formation of the specimens of the so-called Diaseris did not strike 
his attention. 
Specimens in the Challenger collection and in the collection of the British Museum 
show well the different shapes and degrees of growth of the forms ; and in the Challenger 
collection occur two perfect specimens of Cycloseris (Diaseris) freycineti. 
Owing to the thinness and delicacy of the small and large specimens of the greater 
number of the recent species of Cycloseris, and especially owing to the thinness of the 
wall, such specimens would be very easily broken, whether by the action of various 
animals or by the play of winds and waves; a liability which is much increased by the 
fact that these delicate corals are not attached—at any rate not except, perhaps, in their 
very young stages. And this seems to me sufficient to account for the fact that so large 
a proportion of distorted and mended or broken specimens are found in comparison with 
the perfect specimens. 
Such reparation of injury is very common among the Fungide ; and injury to the 
