OP THE MEKGUr ARCHIPELAGO. 



227 



of the outer end scattered up and down it without regular order. 

 Set vertically to the surface, about one layer deep ; size up to 

 about "14- by '052 to '06 millim. Spicules all colourless and 

 transparent. 



Hah. Sullivan Island, sublittoral. 



Two specimens of this species came under my notice for 

 description. The largest, when perfect, must have been about 

 3 inches in height and greatest diameter respectively, stem just 

 below pileus about 2 by f inch in diameter and If in height, erect 

 lobes up to about 1 inch high, marginal folds about 3 to 6 mm. 

 thick. A younger and less perfect specimen differs in the in- 

 ferior prominence of the autozooids and in the dirty white 

 colour. A slight colour ring indicated on the lower part of the 

 stem in both specimens is perhaps due to staining by immersion 

 of this part in mud. The species, from its prominent autozooids, 

 has a considerable resemblance to a low-calicled Madrepora. I 

 see no hint of such a character in either of the two described 

 forms obviously not distantly related to it — L. (Sarcophytwm, 

 Klunzinger) paucijlorum, Ehrb. (Red Sea), and its var. validum, 

 Yon Marenzeller (Andaman Islands, Tonga Islands). It differs 

 from the typical form of this species, but agrees with the variety, 

 in its projecting superior lobes and in the general form of the 

 colony. Prom var. valida it differs in the greater definiteness 

 of the form of the spicules, i. e. in the arrangement of their 

 tubercles in whorls ; the deep-lying spicule of the stem is par- 

 ticularly characteristic, with its wide median space, well-defined 

 ends, and heavy compact tubercular whorls ; the proportions, 

 generally speaking, of the spicules closely resemble those of var. 

 valida. 



Spongodes, Milne- Edwards Sf Haime. 



Distinct as species of Spongodes commonly look at first sight, 

 it is not so easy to feel positive about specific distinctions. 

 Colour, however, appears to furnish fairly constant characters. 

 Of spicular characters, those of the characteristic "projecting" 

 spicules (as I have below designated those peculiar to the genus), 

 and of the spicules of the stem, seem to give the best promise, 

 especially the condition of the outer end of the former and the 

 shape and the tubercles of the latter. Dr. Gray's character of 

 the arrangement of the zooids on the lobules separates the genus 

 very conveniently into groups. With regard to distribution, 



i i . 



