1909.] MARINE FAUNA FROM MERGUI ARCHIPELAGO. 907 



typical figure-of-8 form of two individuals in the process of 

 dividing. The joining of the septa of the fifth cycle over those of 

 the fourth is decidedly marked. At the point where the junction 

 takes place the single septum formed by the union of the two 

 septa of the fifth cycle becomes strongly convex, as Milne-Edward.s 

 has already pointed out, and the lower half projects inwards 

 towards the columella beyond the septa of the first three cycles. 



Localities. — Station VI. Near Grant Island. Bottom: rock and 

 sand oi- rock and mud. Depth : 3-7 fathoms. 14 specimens. 



Station XXX. Fly Island. Bottom : rock and sand. Depth : 

 8-15 fathoms. 307 specimens. 



Station XXXIII. Christmas Island. Bottom: rock, sand, a,nd 

 mud. Depth: 8-23 fathoms. 1 specimen. 



Genus Dendrophyllia Milne-Edwards & Haime. 



In the present collection there are three species of Colonial 

 Madreporaria which have points in common with Milne-Edwards 

 and Haime's original three genera Dendrophyllia, Ccenopsammia, 

 and Lobopsammia, and the diificulty of deciding to which they 

 belong is such that it will probably be convenient in the future to 

 recognize but one genus instead of three. In the original diagnosis 

 of the genus C(enopsam/mia Vjy Milne-Edwards and Haime, the close 

 connection between the three genera was pointed out ; but Beu- 

 drophyllia and Lobopsammia are distinguished from the former by 

 the star-like appearance of their calices (I'aspect etoile des calices). 

 Klunzinger has remarked on the same difference ; speaking of 

 Coe.nopsammia he says : — " Daher ist hier aiich das fiir die Familie 

 charakteristische Zusammenlaufen der kleineren Septa hier nicht 

 oder wenig ausgesprochen (zum Unterschied von Dendrophyllia)'^ 



Copnopsammia has only three complete cycles of septa ; Dendro- 

 phyllia and Lobopsammia have four complete cycles. Lobopsammia 

 pr-opagates by fissiparity ; both Dendrophyllia and Coenopsammia 

 propagate by gemmation. Verrill has already placed Coeno- 

 psammia with Dendrophyllia, and in this has been followed by 

 most subsequent authoi-s, and the specimens in this collection not 

 only justify this conclusion, but make it seem advisable to include 

 Jjobopsammia in one and the same genus with Coenopsammia and 

 Dendrophyllia. The species I identify as Dendrophyllia coGcinea 

 {Ccenojysammia coccinea M.-Edw. & II.)in general appearance and 

 mode of growth is very like C. tenuilameUosa and C. ehrenbergiana 

 {M.-Edw. & H. [25. pi. i. figs. 11 & 12]), and Wayland Yaughan's 

 figure (37. pi. xlvi. figs. 6 k 6 «) of Verrill's original specimen of 

 Dendrophyllia manni [Coenojysammia m,anni Verrill). The septal 

 arrangement is very dendrophyllid in character, the star-like 

 appearance caused by the union of the fourth to the third cycle 

 of septa about halfway between the lip of the calice and the 

 columella being pronounced. The septa are so irregular as regards 

 V)oth cycles and systems that they aflbrd no certain guide. In 

 the smaller colony of the East African species there is one 

 individual with a complete fourth cycle {Dendrophyllia and Ijobo- 



