736 MR. J. STANLEY GARDINEB, ON [June 6, 



1'1-1'9 cm. in diameter. The costse of the first two cycles of 

 septa are distinct down the whole of the corallum, but those of 

 the third cycle extend only for about 3 mm. and are very small. 



The septa vary up to 2 mm. in exsei'tness and are entire, rounded 

 at the top, sloping down rapidly to the axial fossa, which has no 

 columella, being closed in below by trabeculse from the septa. 



In the living colony there are often two or three mouths, where 

 there is no trace of more than one system in the corallum beneath. 

 The polyps, when fully expanded, rise for about 3 cm. above the 

 skeleton, but contract to less than 1 cm. in spirit ; their tentacles 

 are about 2 cm. long and not retrusible. The species is exceedingly 

 difficult to preserve on account of the enormous quantity of mucus 

 secreted, the least touch with the hand under water causing the 

 complete obliteration of the stomodseum and peristome. The 

 colour of the polyps varies from green-yellow to olive-brown, 

 always markedly greener towards the stomodsea. 



Rotuma ; in pools of outer reef, very local in distribution, often 

 forming large masses, 3-4 feet across. 



I have been unable to determine the method by which the calices 

 are divided, but in elongated corallites two of the larger septa 

 sometimes fuse across the fossa and form a division. The thecal 

 walls of the two sides then grow horizontally over the edges of 

 these and fuse, flattening out to a breadth of about 2 mm. The 

 young corallites next broaden out in a plane at right angles to the 

 elongation of their parent corallite, themselves later to undergo a 

 similar division. 



Genus Mussa. 



Mussa (pai's), Oken, Lehrb. der Naturg. i. p. 73 (1815). 



Musm, Dana, Zooph. p. 173 (1848). 



Mussa, Milne-Edwards & Haime, Cor. ii. p. 328 (1857). 



Although I have examined a very large number of specimens of 

 this genus both in the British and Cambridge Museums, I have 

 completely failed to find any characters of good general specific 

 value. Tet the specimens seemed to fall naturally into from eight 

 to ten groups, none of which, however, were sharply marked off 

 from their neighbours. The characters of the several septa, on 

 which Dana very largely relied, do not seem to me to be of any 

 great importance, and not uncommonly in the same colony septa 

 corresponding to 4 or 5 of Dana's species can be found. The 

 depth and shape of the corallites, too, depend very largely on the 

 position of growth, i. e. whether the colony has plenty of room or 

 is crowded with other organisms. Mainly depending on these 

 characters Milne-Edwards and Haime described no less than 

 thirteen species, most of which will indubitably be found to be 

 synonyms. 



Although I depend partially on the general form of growth, 

 yet the more important characters seem to me to be — the extent 

 and form of the epitheca, presence or absence and spinulation of 

 costse, and characters of the columella if present. The arrangement 



