13 



pillae; and either 1 or 2 adambulacral spinelets may occur in the same 

 species (e. g., tenebrarius). I do not agree with Ludwig who considers 

 the segmental papillae as of primary importance in dividing Porcellan- 

 aster, and the cribriform organs as of no importance. The latter are 

 absolutely constant, while the segmental papillae are variable in number. 

 From an examination of 51 specimens of E. tenehrarius it is evident 

 that the range of variation is much greater than I supposed in 1905 

 where I had only 6 specimens. For instance, while there are always 

 3 cribriform organs, the number of segmental papillae, that are easily 

 recognized as such, ranges from about 5 to 15 on either side of a fur- 

 row. The superomarginal spines are often entirely lacking, or may be 

 absent from first 2 or 3 plates , or scaltered here and there almy ray. 

 The segmental papilla on outer part of ray is transformed into an or- 

 dinary spinelet ; and it is evident that the segmental papilla of Eremi- 

 caster and the aboral adambulacral spinelet of Porcellanaster are homo- 

 logous structures. Consequently the outer part of the ray in typical 

 Ei'emicaster has the same adambulacral armature as the whole ray of 

 Porcellanaster. It is but a slight step to the total disappearance of seg- 

 mental papillae in Eremicaster pacificus (Ludwig). The extent of the 

 »outer part of the ray« varies considerably in different specimens for it 

 is not always evident where a papilla ceases to be such and become a 

 spinelet. I have examined specimens of pacificus (which lacke seg- 

 mental papillae but has 3 cribriform organs) and the species appears 

 to be much more closely related to E. tenebrariiis than to P. caeruleus. 

 The reverse would naturally be true if the species belongs in Porcellanaster. 



I have carefully compared the figures and description of Eremi- 

 caster ivaltharii (Ludwig) with my E. tenehrarius. I have no doubt that 

 they are the same species, for the only tangible difference — the pre- 

 sence of 2 adambulacral spinelets in tenebrarius — is not constant, 

 many specimens having either 1 or 2, or only 1. In fact there is more 

 difference between the extremes of 45 specimens from a single station 

 off southern Alaska than exists between typical tenebrarius from Cali- 

 fornia and ivaltharii of the Panama region. The former name has 

 about a mouth priority. The range of E. tenehrarius is thus southern 

 Alaska to the Galapagos Islands, and 1569 to 2259 fatlioms. 



Eremicaster pacificus is found in Bering Sea, south of the Pribilof 

 Islands, in 1771 fathoms. Its range is Bering Sea to the equator, in the 

 eastern Pacific. 



In raising Eremicaster to the rank of a genus Prof. Ludwig, 

 curiously enough, changed the type [tenehrarius] to crassus Sladen. This 

 is contrary to a fixed law in nomenclature, Eremicaster whether genus or 

 subgenus, will stand or fall with tenehrarius as type. The two species 



