42 VERRILL 



Sladen and Perrier disagree both as to the character and limits of 

 the two groups, because certain genera and species present characters 

 of both groups, or have intermediate characters. Therefore it 

 seems necessary to reunite them in one family. However, it may 

 be convenient to retain the groups as subfamilies or sections with no 

 very definite limits. 



A pseudostichasterial condition is sometimes produced by the 

 existence of clusters of spinules in regular rows, on the transverse 

 dorsal connective ossicles. This cannot always be determined with- 

 out removing the investing integument. 



Perrier considered the characters of the dorsal spinules as of para- 

 mount importance in this group. He would place all the genera with 

 distinct spinules or spines in Asteriidae, and those with granule-like 

 spinules in Stichasteridas ; ' but in fact there are all intermediate 

 gradations in the character of the armature, and it is often difficult 

 to say whether these structures, in some species, should be called 

 " granules " or " spinules." This seems to me a character only of 

 specific value.' 



Subfamily ASTERIINJE. 



Asteriida (restr.) Sladen, op. cit., 430, 560, 1889. Perrier, 1894, pp. 105, 

 128; 1896, pp. 25-31. 



As here limited this subfamily corresponds nearly with the re- 

 stricted family Asteriidse in the systems of Sladen and of Perrier. 



The Asteriinag ' are chiefly characterized by the more or less 

 openly reticulated arrangement of the dorsal and lateral ossicles, 

 especially of the dorsal ossicles, which are usually more stellate or 

 lobulate and more slender than in the Stichaster group, and generally 

 united either by the ends of the lobes or apophyses, or else by means 

 of interpolated ossicles, thus allowing more flexibility in the rays, and 

 larger papular areas. 



The Stichasterinae, on the contrary, are chiefly characterized by the 

 more or less regular subtesselated or imbricated arrangement of the 



' 5. gracilis and 5'. albula, with otiier related forms, have definite spines. 



'The genus Calvasterias Perrier, 1895, has about five series of lobed and 

 imbricated, flat dorsal plates, bearing few small short spines, and entirely 

 covered with a thick, nal<ed, soft, canaliculated dermis. 



The character of the plating should cause it to be placed in the Sti- 

 chasterinae. 



It is monacanthid and there is a row of small spiniferous, interactinal plates, 

 at least in C. stolidota Sla., of Chile and the Falkland Islands. 



"This form of spelling the word seems necessary, because it is derived from 

 Asterias, not from Aster. 



