ccenenchyma, and their position is marked by a slightly elevated wart-like 

 projection with a central pore, around which may be discerned an eight- 

 rayed figure. 



The spicules of the ccenenchyma (Fig. 4, a, h, and c) consist almost entirely 

 of triple-stars. Of these there are two distinct types — (a) triple-stars with 

 very densely-warted and closely-set heads, and with very short handles ; 

 (6) triple-stars with much longer handles and with the heads covered with 

 openly-set long warts ; (c) a few double clubs also occur. Most of these spicules 

 are red in colour. The spicules of the axis are colourless ; they are very 

 irregular in shape. Some are almost smooth, while others are warty. A 

 few of the characteristic variations in these spicules are shown in Fig. 4 (d-z). 



The following are some of the measurements ^ in millimetres : — 



1. Ccenenchyma (Fig. 4, «-c). 



(a) Triple-stars with short shafts and densely-warted star 

 portions. 



0-076 ; 0-068 ; 0-065 ; 0-038 ; 0-03. 



(6) Triple-stars with long shafts and witli few simple waits 

 in the star portion. 



0-087 ; 0-068 : 0-0C5 ; OO.nS. 



(c) Double-stars — a few of these occur, but they are usually 

 small and may be undifferentiated triple-stars. 

 0072 X 0-03 ; 0057 x 0-042 ; 0038 x 0019. 



2. Axis (Fig. 4, d-i). 



0-152x0-076; 0-114x0065; 0-106x006; 0106x0075. 



There can be no doubt that our specimens belong (1) to the Order 

 Pseuda.xonia, and (2) to the Family Sclerogorgidre, which is thus defined by 

 Bourne (Lankester's Treatise of Zoology, part ii., "The Anthozoa," p. 25): 

 "The medullary mass forms a distinct axis consisting of closely packed 

 elongate spicules with dense horny sheets. The axis does not contain 

 solenia but is surrounded by longitudinal canals, i.e., by large solenia which 

 are connected with tiie zooid cavities by smaller ramifying solenia." 



The position of the Gorgonellidffi, in general classification, is a matter 

 of some difficulty, but in a memoir on that family about to be published, 

 we have suggested that forms such as the one under consideration may be 

 annectent between types like Suhcroyor/jia and types like Jinicella, but our 

 present knowledge does not warrant the inclusion of this form in the 



I The measurement of the triple-stars taken is the maximum, i.e., from the extreme 

 end of one " star " to the extreme end of an .id jacent " stjir." 



