they are not irifid, as is commonly the case in this species, 

 but end irregularly; they are all covered with a thick skin, 

 which often unites with the skinfof the neighbouring spines, 

 forming thus a thin web along the frames. But one or two 

 papulae in each mesh. fPl. 11. Fig. 7). 



For comparison a part of the abactinal skeleton of a typical 

 Cr. smiguinolenta is figured (PI. II. Fig. 9). The difference is very 

 striking, but as the species is so extremely variable, transitional 

 forms may be found. In Fig. 8 is represented a part of the 

 abactinal skeleton of a specimen from the Cattegat. Here 

 the spines stand only 3 — 4 together, and there may be one or 

 a few small plates in the meshes, some of them even bearing 

 a single spine. These are not trifid and not covered by 

 thick skin. 



The adambulacral spines do not offer any marked difference 

 from those of the typical Cr. sanguinolenta; there is only a 

 single series of spines outside these, whereas there is a double 

 series in the typical form. The inner small spine is present, 

 but often little developed. — The arms are not very slender; 

 R = c. 35 mm., width at the basis с 10 mm. (The measures 

 cannot be quite exact because of the bad preservation of the 

 specimens). 



These specimens are very interesting, showing how close 

 is the relationship between the genera Cribrella and Echinaster. 

 in fact they might as well be referred to Echinaster., there 

 being, as far as I can see, no distinguishing character of any 

 importance. Especially they resemble Echinaster scrobiculatus 

 Dan. Kor. so much (the spines of this species being small and 

 arranged in a similar manner as here, even occasionally two 

 together. Danielssen & Koren: Op. cit. p. 40) that it seems 

 rather irrational to refer them to two different genera. Also M. 

 Sars (Norges Ech. p. 86) has found specimens of Cr. sanguino- 

 lenta with the spines arranged in the manner here described. 



