OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 249 



The upper and under arm plates have the forms which I have gave 

 for them in my original description, the former with the proximal 

 angle truncated and slightly produced and the latter with the distal 

 border concave. The under arm plates, while broader than long, are 

 perhaps a little less broadened than in my type; this comes- about 

 doubtless from the difference in size of the individuals ; in the speci- 

 men which I figured in 1904 ('04a, -fig. 76), the diameter of the disk is 

 10 mm., while in the Albatross specimen this diameter does not exceed 

 6 mm. The arm spines are flattened, rather thick and strong, slightly 

 broadened at the tip; they are finely denticulate throughout their 

 whole length. 



The hook formed by the first ventral arm spine is fairly well de- 

 veloped (pi. 100, fig. 4:c) ; it usually shows three branches, sometimes 

 only two, the terminal branch very much stouter than the two others. 



The tentacle scale, which is of medium size, is oval, ending in a 

 stout point often accompanied bv a few others very much smaller 



The general color is an olivaceous gray on the dorsal surface of the 

 disk ; the arms are lighter, with traces of reddish bands. The ventral 

 surface of the disk is light gray and the arms show darker annula- 

 tions; there is not the least indication of a lighter or darker longi- 

 tudinal line on the arms. 



In 1904 I mentioned that the specimens upon which I based this 



species, and which had been labeled by Marenzeller 0. stelligera, 



were in reality very different from the latter ; I had not then had the 



opportunity of studying O. koreana, and I was not able consequently 



to compare it with 0. marenzelleri. I now maintain that these two 



species are rather closely related, and I notice especially a great re- 



semblance between O. marenzelleri and 0. hylodes from Japan, which 



H. L. Clark described under a different specific name, though saying 



that he was tempted to consider it merely as a variation of 0. koreana. 



The only difference which I can find is shown by the spines which 



cover the dorsal surface of the disk in O. hylodes and which pass 



over onto the radial shields, where they become less closely crowded. 



These spines differ from those which I find in O. marenzelleri, 



which, moreover, had perhaps better be called club spines. They have 



a smooth surface, and they bear only two or three unequal spinules 



at their tip ; they appear to me to be shorter and less stout than in 



O. hylodes, while on the other hand the arm spines are more slender 



in the latter. 



Ophiothrix marenzelleri also shows affinities with O. caespitosa 

 from Australia ; to facilitate comparison I include three photographs 

 of a specimen which I have of this latter species from Port Jackson, 

 Lyman's type locality (pi. 38, figs. 6, 7, 8). 



