CLARK: BRITTLE-STARS. 335 



to the oral shields are red. Rarely the terminal half of the arms shows 

 indications of three or four transverse dusky bands; these are most 

 evident in the wholly green specimen. In the wholly red specimen, 

 the disk and arms are of nearly the same shade, a rosier red than 

 usual and the arms are slightly blotched and are tipped with reddish 

 white. Preserved specimens, whether alcoholic or dry hold their 

 colors fairly well, but of course they are duller than in life; the greens 

 seem to change more than the reds. 



This is a very handsome and noteworthy species, quite different 

 from any of its West Indian congeners. In many structural features, 

 it resembles 0. panamcnsis but aside from the striking and important 

 color-differences, the granulation of the disk is very much coarser and 

 the arms are shorter and stouter. It occurs commonly on those 

 parts of Buccoo Reef which are accessible only at the lowest tides, 

 its usual companions being Ophiodcrma cincrcum, 0. guttahwi, and 

 Ophiomyxa flaccida. It was only found elsewhere on two or three 

 occasions. 



Ophioderma squamosissimum. 



Liitken, 1856. Vid. med., p. 8. 1859. Add. ad hist. Oph., pt. 2, pi. 1, fig. 

 7a, 76. 



Plate 4, fig. 1; Plate 6, fig. 3, 4. 



That this superb ophiuran is one of the handsomest and most bril- 

 liantly colored of marine animals will not be questioned by any one 

 who sees it alive. Hitherto known only from the type at Copenhagen, 

 from an unknown locality in the West Indies, its discovery at Tobago 

 was one of the notable features of our stay there. We found it only 

 on that part of Buccoo Reef, which was accessible at the lowest tides, 

 and even there it was very rare; only five specimens were seen and 

 on only one day were two specimens secured. These are all smaller 

 than the holotype and range from 11 to 22 mm. in disk-diameter. 

 They agreed exactly in color, which was a brilliant red, about the 

 shade designated as scarlet by Ridgway (1912), or very near 

 Klincksieck's and Valette's no. 76. The disk was slightly marked 

 with a somewhat yellower shade. Unfortunately the color bleaches 

 very rapidly and completely in alcohol (the holotype is recorded 

 as having the color "entirely bleached out") but it is appar- 

 ently not affected by formalin. Specimens treated with formalin 

 and corrosive sublimate, and then dried, retained the color well. for 



