DENDY— HOMOSCLEROPHORA AND ASTROTETRAXONIDA 269 



texture is hard and tough. The colour of the upper surface is rather dark brown, 

 internally much paler, with, to the naked eye, a sharp boundary line between cortex and 

 choanosome. Under a pocket-lens the huge spherasters can be distinctly seen both on 

 the surface and in the interior (when cut) ; they give the surface a characteristically 

 harsh feel. 



The cortex is about 0'3 mm. thick and contains a great many of the large spherasters, 

 for the most part arranged in a single layer at the surface. It is fibrous, and around the 

 spherasters the fibres are concentrically arranged. The dark brown pigment granules are 

 most abundant in the outermost part of the choanosome, just beneath the cortex, but 

 occur also both in the choanosome below and in the cortex above this layer. Numerous 

 narrow inhalant canals (pore-canals) penetrate the cortex vertically, each starting above 

 from a single dermal pore. These canals are rendered conspicuous by the numerous 

 pigment granules that accompany them. Probably they unite in groups on their way 

 through the cortex as in other species of Chondrilla. 



The skeleton consists of spherasters (Plate 48, figs. 8 a, 8 6, 8 c) only, varying much in 

 size and in the shape of the rays. They are abundantly scattered in the choanosome as 

 well as in the cortex. The fully grown ones, measuring about 0"14 mm. in diameter, are 

 by far the most abundant. There appears to be no difierence between the cortical and 

 choanosomal spicules, much the same range of variations occurring in both situations. 

 The small ones are probably young forms ; I have measured one with a diameter of no 

 more than about 0'04 mm. ; they have very numerous, simple, smooth, conical rays, 

 springing from an enormous centrum, the rays only about 0"004 mm. long. The large 

 ones are of two principal kinds, with intermediates : — (1) with sharp-pointed conical rays, 

 somewhat inflated towards the base ; often slightly roughened except for the tips of the 

 rays (fig. 8 a) ; (2) with the rays truncated and roughened at the ends (fig. 8 h), or even 

 reduced to short, subcylindrical projections with roughened ends (fig. 8 c). The inter- 

 mediate forms show all degrees of truncation and roughening of the rays*. 



I have already, in discussing the genus Auroral, given my reasons for believing that 

 Carter's Chondrilla sacciformis is a true Chondrilla, and I need not repeat them here. 

 The examination of a type specimen in his cabinet has convinced me that the " Sealark " 

 specimen is specifically identical, and there appears little doubt that this is also the case 

 with Thiele's Chondrilla grandistellata. I hardly know what Thiele means by the 

 statement that a distinctly differentiated cortex is not present in his specimens. At the 

 same time it is true that the line of demarcation between cortex and choanosome is by no 

 means everywhere clearly defined in mine, possibly owing to the indifferent histological 

 condition of the material. 



The enormous size of the spherasters is very remarkable in comparison with such 

 species as C. australiensis, C. mixta, &c. 



Previously known Distribution. Mauritius (Carter) ; Ternate (Thiele). 



Register No., Locality, &c. xi. 4, Saya de Malha, 7.9.05, C. 19, 29 fathoms. 



* The form represented in Fig. 8 c is probably alone fully grown, all the others being developmental 

 stages. 



t Vide p. 245. 



