226 K. KINOSHITA : ON THE KEROEII>Il>.E, A NEW FAMILY OF 



The Kerocides specimens at my disposal show, at the first sight, 

 such a diversity among themselves that they seem to be separable 

 into several forms or species. A close examination, however, has led 

 me to accept the view that the differences are not strong enough to 

 be employed for specific distinction. 



Without going into the systematic status of the specimens, I shall, 

 in this paper, content myself with giving a note on the structure of 

 the axis, the unique character of which is entirely unknown in the 

 known Gorgonians. 



The axis is cylindrical in form, not jointed ; it runs centrally 

 through the stem and branches. It is rigid and brittle except in the 

 terminal parts which are soft and flexible. The colour is yellowish 

 or dark brown, owing to that of the horny matrix. In all the cases 

 examined, the spicules participating in the construction of the axis, 

 are not at all colored. 



The axis consists of two parts, viz., a central cord and a cortical 

 layer (respectively the Centralstrang and the Rindenlage of Kölliker). 



The central cord is structurally exactly the same as that of true 

 Gorgonians (figs, i and 2). The cortical layer is composed, as already 

 described by Wright and Studer, of smooth spindle-like spicules ar- 

 ranged longitudinally and of a fibrous horny matrix, which tightly 

 binds up the spicules (fig. 4). The latter sometimes show an arrange- 

 ment in concentric rings. The horny substance of the axis runs out 

 into the hyaline mesoglcea without any distinct contour. The axis is 

 surrounded by wide and narrow solenia, which take no regular ar- 

 rangement. 



Tracing the axis towards the tip of the twigs, it comes to lie in 

 the partition between the two terminal polyps which are placed usu- 

 ally nearly opposite to each other, and ends beneath the superficial 

 spicule layer of the very tip (fig. 1). For some variable length the 

 axis is composed of the central cord only and is entirely devoid of 

 the cortical layer. The central cord, so far as it remains naked, is 



