90 s. goto: 



beautifully ou sbiiuiug the skiu. Tlie "svall of the stomach is full of small, 

 irregular spécules (PI. XYI. Fig. 4). 



*' lu one of the iuteiTaclia I foimd ou this specimen a cluster of eggs, 

 attached to oue of tlie genital openings. The eggs are rather large, O.G-0.7 

 mm in diameter. The ovaries are seen to contain eggs in all stages of 

 development, from quite yoimg to such as are quite matm-e, ready to be 

 laid. These facts make it very probable that this species has not i)elagic 

 larvœ ; evidently the eggs must be laid at different epochs, since they do 

 not become ripe at the same time, as is generally the case in those forms, 

 where the eggs develop into pelagic larvœ. 



" The present sj)ecimen differs not inconsiderably from the typical form 

 in the marginal plates being somewliat spinous and not so regular in shape 

 (PL XrV", Fig. 4, to comj^are with PI. XFV, Fig. 6, representing the cor- 

 responding plates from a specimen of the typical form). The limit between 

 the upper and lower marginal plates is indistinct, on account of a thick 

 covering skin. Wliether tliis specimen represents a distinct variety or perhaps 

 an individual abnormality can, of course, not be settled on the evidence of the 

 present material. But I am inclined to regard it as an abnormal specimen. 

 In a specimen from the Kara Sea I have foinid traces of a similar in-egularity 

 in the marginal i)lates, but much less developed than in the present specimen. — 

 Otherwise it agrees with ICauschewskij's cold water variet}-. 



" Concerning the most interesting geographical distribution of tliis 

 species, see below, p. 294." The wide distribution of this and several other 

 species of starfishes is cited as an argument in favom- of A. H. Clakk's 

 " Polar-Pacific " region. 



FiSHEE adds to the Hst of synonyms the Ctenodiscîis cmstralis of Lovén 

 MS. LÜTKEN, a point on which I would not express my own opinion without 

 examining the original specimens. He then gives a very exhaustive account 

 of G. crispat'us [Fishek, :11, p. 31] : 



"Diagnosis. — Kays five, exceptionally four or six. Pt=:1.66 r to 3.16 r. 

 General form stelkito-pentagonal, to stellate, extremely variable. Abactinal 

 surface more or less tumid and usually' with an elevated cone in centre 

 of disk ; paxilla) variable, usually low, with few to many spinelets which 



