128 BULLETIN 76, UNITED STATES NATIONAL MUSEUM 



broader than high, especially in those of the supramarginal and intermarginal chan- 

 nels. The distal margin is finely denticulate and is often slightly produced in the 

 middle. This is a prolongation of a low median ridge or carina which is sometimes 

 found on the outer face of the valve. 



Amchitka Island: This is one of the Rat Islands between the Andreanof Group 

 and the Near Islands ( Attu and Agattu). The smaller of the five specimens resemble 

 forma shumaginensis. The largest (R 38 mm.) has a more even surface than the 

 Atka variety and also resembles the sparser spined intermediate specimens from Attu 

 and Agattu. The bivalved pedicellariae are mostly smaller than typical with jaws 

 usually higher than wide on outer half of ray. 



Attu and Agattu: The specimens from Agattu (pi. 56, fig. 3), the largest of 

 which has R 65 mm., r 18.5 mm., are rather more uniform in appearance than those 

 from Attu (fig. 4). They represent a recognizable subrace of alaskensis notable for 

 the robust abactinal spinelets which are capitate, subtruncate, and do not define the 

 papular areas in reticulate pattern, but tend to form groups which along the carinal 

 line arc crowded and sometimes a trifle elevated (especially in young specimens). 

 These groups occur also on the dorsolateral region but with a few exceptions are not 

 so well defined. The straight pedicellariae are about intermediate between those of 

 aJaskt nsis and those of asiatica. 



Numerous young specimens from Agattu (R 6 mm. to 25 mm.) have the carinal 

 groups of two or three spinelets relatively more prominent than in large examples, 

 while on the dorsolateral area the spinelets, of two or three sizes, are rather widely 

 scattered. The lateral pedicellariae are thick-ovate with straight-sided or slightly 

 tapered jaws, not greatly unlike the larger sort found in asiatica. 



The specimens from Attu (pi. 56, fig. 4) agree with those of Atka in having 

 considerable disparity in size between major and minor abactinal spinelets, and in 

 having the former in prominent, though small, groups which give a very uneven 

 appearance to the abactinal surface. This is simply an exaggeration of the condition 

 in Agattu examples. Four specimens are practically the same as those from Atka. 

 The others have rather coarser spinelets. The bivalved lateral pedicellariae are rather 

 Larger than in the Agattu examples, but the jaws are not so wide as in Adakh speci- 

 mens, that is, are not wider than high. 



There seems to be in this region (the Near Islands) a fairly definite subrace 

 intermediate between alaskensis and asiatica, as exemplified by the majority of 

 specimens from Agattu and some of those from Attu. Then there are (especially 

 at Attu) a number of variants which may be the result of crossing of alaslcensis with 

 iispar. One evidence of this is the instability in the size of major pedicellariae; 

 another the formation of prominent groups of carina! spinelets and the dwindling in 

 size and number of the dorsolateral spinelets. 



Forma SHUMAGINENSIS, new forma 



Plate 48, Figure 3, 3a; Plate 57, Figure 1 



These specimens differ from large, well-developed Unalaskan examples in having 

 slenderer, cylindrical or slightly tapered spinelets which are only occasionally sub- 

 capitate. They stand not very close together on the skeletal ridges and do not form 



