122 BULLETIN 76, UNITED STATES NATIONAL MUSEUM 



lariae gray or greenish gray, and straight pedicellariae whitish or yellowish white; 

 abactinal and marginal spines bright vermilion on distal half or at tip, shading rather 

 quickly into dark blue or purple at the base; adambulacral and actinal spines lighter, 

 with only a little blue at base; tube-feet rich cadmium yellow, the disk pale straw 

 yellow. 



Variations. — The principal variations have been pointed out in the foregoing 

 description. In order to eliminate the question of geographic variations, 10 speci- 

 mens have been selected at random from the ample La Jolla collection. These range 

 in size from R 49 to R 81 mm., being mostly over R 65 mm. 



Four have the dorsal spines frankly in three series and six have extra spines 

 laterally to the dorsolateral series (a second incipient dorsolateral series). There is 

 no age (size) correlation with the abundance of larger straight pedicellariae, nor with 

 the presence of actinal spines (for all 10 have them in some degree), nor with the num- 

 ber of distal adambulacral plates having only one spine (the extremes are present in 

 both groups). 



If this sample selection of 10 specimens is divided into two groups on the basis 

 of the number of actinal spines: (a) six specimens have 10 or more spines, and (6) 

 four have less than 10. There is no correlation with size, group a ranging from R 55 

 to R 85 mm., and group b, from R 49 to R 76 mm. It is true, however, that most 

 specimens of small size have only a very few actinal spines, or else lack them; large 

 specimens likewise may have only a few spines. In group a, a specimen with a 

 radius of 64 mm. has 14 actinal spines while another with radius of 51 mm. lacks them 

 entirely, and a third of 54 mm. has 11. This will indicate the haphazard occurrence 

 of this character. In group a, three specimens have three rows of dorsal spines and 

 three have extra laterals; in group b, one (the smallest) has three series of dorsals, 

 and three have extra dorsolateral spines. In group a, five have the large straight 

 pedicellariae abundant enough to be easily seen at a glance, and in one they are scarce; 

 in group b, three have abundant pedicellariae, and one (the smallest specimen, R 49 

 mm.) has only two or three incompletely developed large pedicellariae. Thus if 

 two groups were made on the abundance of pedicellariae eight would have the pedi- 

 cellariae well-developed and two poorly developed. Small examples generally have 

 very few, or even none, of the large pedicellariae; yet one of the above two specimens 

 has R 70 mm., which is well beyond the medium size for this species. 



Three of the sample 10 have mostly only one spine on each adambulacral plate 

 of the terminal third of the ray, while on the other seven the terminal area is more 

 restricted. 



There is some variation in respect to the stoutness of the ray, but the size of the 

 gonads will affect this feature. 



A specimen from San Francisquito Bay, Lower California, {Albatross, 1911) 

 has R 53 mm. and 15 actinal spines, extending half the length of ray. The dorsal 

 spines are a shade slenderer than in the average northern specimens, rather sharp, 

 and largely covered by the pedicellaria sheaths which are not retracted. The large 

 straight pedicellariae are very few and less heavy than in La Jolla examples, except 

 along the furrow margin where they are numerous (but not in Gulf of California 

 specimens, No. 1214, M. C. Z.). The adambulacral spines are a trifle slenderer and 

 longer than in most northern specimens, especially proximally, but the Gulf of 

 California specimens are not tangibly different from small La Jolla examples. This 



