50 1UI.I.KT1N 76, UNITED STATES NATIONAL MUSEUM 



In Greenland, Spitzbergen, and Kara Sea small examples the adambulacral 

 Bpinea normally lack pedicellariae (where in Bering Sea specimens the outer spines 

 would have small clusters). In large specimens the outer spine of diplacanthid 

 plates may carrj a conspicuous cluster of pedicellariae, but not the furrow spine. 

 Only a few specimens are large however; most of them lack adambulacral spine 

 pedicellariae, excepl sporadically. 



A considerable number of dwarf specimens are among those recorded in the list 

 of localities, especially from Indian Point, Bering Strait (3582), Bering Strait (16591), 

 and Plover Bay, It) to 25 fathoms (16584). Duplicate specimens of the last two 

 lots were examined by VerriU and are recorded (1914, p. 123) as L. arctica. The 

 Plover Baj specimens are in poor condition. They are all small, with tumid rays 

 and numerous short spinelets. Some of them carry eggs in the stomach, a habit so 

 far as known, not characteristic of arctica. The adambulacral plates are mostly 

 diplacanthid, and the marginal spir.es have complete rings of pedicellariae. The 

 smallest specimen known to be carrying eggs has R only 12.5 mm., r 6 mm. 



Young.- --Very small specimens have that generalized appearance which renders 

 identification, in the absence of adults, extremely unreliable. It is difficult, in a 

 species which breeds when small (K 12.5 mm.) to differentiate between real young and 

 dwarfed specimens. There is a very real difference, in that the dwarfed specimens 

 have more spines. They remind one of dwarfed trees that have managed to add new 

 lings without much increase in size. Such are numerous small examples (some 

 carrying eggs) from Plover Bay, Siberia, and two examples from Indian Point, Bering 

 Strait, 17 fathoms (No. 3582), the largest of which has R only 13 mm., but which 

 resembles an adult of forma cribraria in miniature. 



Undoubted young, with R 7 mm., have a 4-faced ray, with gently convex abacti- 

 nal surface. The abactinal spinelets, five or six across the abactinal area, already 

 are arranged in transverse rows, with a pretty definite longitudinal arrangement, in 

 about five series, in those specimens which are probably destined to become forma 



.laml'ica (station 3252). Of crossed pedicellariae there are a few abactinally and a 

 circle around the marginal spines; and of the straight a few along the edge of the fur- 

 row. There is one scries of supermarginal, one of inferomarginal, and two of adam- 

 bulacral spines. The abactinal plates, relative to width of ray, are very large and the 

 papular areas small. A specimen with R 14 mm. has eight or nine spinelets across 

 the ray near base, two inl'eromarginals on distal half of ray, one to five crossed pedi- 

 cellariae to each abactinal spine, and fairly well developed marginal wreaths. The 

 abactinal spines are quite regular, but T think the specimen was destined to become 

 forma cribraria. A breeding specimen of the same size (forma groerdandica) from 

 Bering Strait \'o. 16591 ) differs in having complete circles of crossed pedicellariae 

 around the abactinal spinelets, mostlj two supermarginal and two inferomarginal 

 spines with thiol wreaths of pedicellariae, and large straight pedicellariae on the 



axils of the rays. 



Anatomical notes. - As mentioned under variations there is a good deal of indi- 

 vidual difference in the abactinal skeleton which is of the very open, irregularly reticu- 

 late type, characterized by slender plates and on the rays by unusually short but 

 broad meshes, one to three of which intervene between the carina! and suoeromarginal 

 scries. The details arc scarcely twice alike. The carina] plates are three or four 



