12 BULLETIN 76, UNITED STATES NATIONAL MUSEUM 



found only al tit.- base of the ray, one and two alternating being the rule on the distal 



' The number of abactinal .Tossed pedicellariae is variable, but for accurate 

 comparison the character is too difficull to handle precisely. The size of the crossed 

 pedicellariae is, however, useful. There is an increase in size from northern Japan 

 to Bering Strait, while in cistern Bering Sea there is a further increase with depth. 

 !„ shallow water a curious variety with heavier skeleton and spines (forma anomala) 

 has Larger pedicellariae than the typical form. A comparison of the sizes and details 

 of pedicellariae is taken up in the following section, but the figures will give the best 

 idea of the \ ariations. 



VARIATIONS OF SPECIMENS BY LOCALITY 



1. Amur {Gvlj oj Tartan/).— This is the cotype, collected by Captain Andrea. 

 A- I have hut the one specimen, no notes can be given on the range of variation at 

 the type locality, hut there is no reason to expect that it is any less than at Petro- 

 pavlovsk. Liit ken. however, has described the types fairly completely, and I shall 

 give a few figures of the spines and pedicellariae. R 115 mm.; rays flattened and 

 distorted. 



The abactinal spines give the impression, at first sight, of being mostly conical 

 and pointed, hut many of them are two or three pointed as Liitken writes. Such 

 spines appear to have a notch or a pit at the tip, especially on the radial region; still 

 others are frankly gouge-shaped at the tip. The abactinal crossed pedicellariae 

 i,w compared to the larger (but still small) straight pedicellariae, which are 

 broadly lanceolate, acute, and about 0.375 mm. long. The abactinal crossed pedi- 

 cellariae are about 0.23 or 0.24 mm. long. (PI. 6, fig. 2.) 



There are eight or nine gouge-shaped superomarginal spines, proximally, and two 

 rows of intermarginaJ spines; inferomarginal spines two, broadly grooved or gouge- 

 shaped; adambulacral spines usually two to a plate, the alternate plates with the 

 characteristic furrow spine. 



The actinal straight pedicellariae are broadly lanceolate, sharp and much larger 

 than the dorsal (about 0.75 mm. long proximally on the intermarginal furrow). 



( <>m|>anng the size of the actinal crossed pedicellariae of the cotype with com- 

 parable examples from Kamchatka, and of rollestoni from Japan, we have the 

 following: 



rollestoni (R 125 mini, Hokkaido, 8 0.17 to 0.23 mm. (PI. 6, fig. 1.) 

 amurensis iK 1 1 5 mm.), Amur, 0.25 to 0.28 mm. (PI. 6, fig. 2a.) 

 amurensis i K 130 mm.), Petropavlovsk, 0.3 to 0.36 mm. (PI. 6, figs. 3, 3a.) 

 The actinal crossed pedicellariae of A. amurensis rollestoni are therefore about 

 three-fifths the length of those of A. amurensis amurensis from Kamchatka. The 

 cotype (Amur) is intermediate, hut rather nearer to the northern specimens. 



2. PetropavlovsJc, shore (PI. 1, fig. 3, pi. 3, figs. 2, 3). — There is a good series 

 ranging from K 12 mm. to R 130 mm. (the last with r 35 mm., R = 3.7 r.). The 



1 Specimens from Mororan on tlic south coast. llariac arc small, 0.18 to 0.21 nim, but rarely larger ones 



wrur which measure 0.2f! mm I ever. In Kamchatkan and Bering Sea examples occa- 



i are found wh! bl; larger than tb Specimens of roUcrfont from farther south on theeast 



ive as 0.2? mm. and as sm il! as IS mm. In typical rollestoni the abactinal spines are 



•outer, and I 



