OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 309 



by the keel is so prominent that Liitken and Mortensen have re- 

 ferred to it in the name chosen by them to designate the species which 

 they believed new. The dorsal surface of the arms is elevated in 

 both species, but the upper arm plates of 0. ambulator are less broad 

 than those of O. fastigatus, and the under arm plates have a more 

 sinuous outline, the middle of their distal border projecting into a 

 small and very prominent lobe coresponding to the ventral keel 

 which does not occur in O. fastigatus. The tentacle scales are very 

 unequal in the two species; in 1896 I wrote that in 0. ambulator they 

 are inserted on the same side of the tentacle pore and are not oppo- 

 site each other ; this is correct only for the pores of the first pairs, for 

 beyond the disk the small inner scale becomes inserted on the under 

 arm plate, and it then occurs opposite the other very much larger 

 scale as is the condition in 0. fastigatus. 



In studying O. ambulator anew, I find on one specimen a feature 

 which I had not noticed in 1896, and which, moreover, only occurs 

 on this one individual. This is the presence on the first arm segments 

 of much developed and very evident ventral pores ; I can easily count 

 10 pairs of them. These pores, quite comparable to those which are 

 known in the genera Ophiarachna, Ophiopeza, etc., are placed on 

 either side of the ventral median keel between the anterior border of 

 the under arm plate and the posterior border of the side arm plates. 

 These pores may be easily seen on the photograph which I give (fig. 

 13). I do not find the least trace of them in the other specimens' 

 which I have reexamined, all of which agree with that of which I 

 give a photograph in figure 9. Liitken and Mortensen did not men- 

 tion these pores in their 0. carinatus. I shall have occasion to de- 

 scribe in Bathypectinura conspicua similar variations, regarding ven- 

 tral pores, certain specimens possessing up to seven or eight pairs, 

 and others completely lacking them ; comparable differences also oc- 

 cur in OpMarachna affwiis. The presence or absence of these proes 

 therefore cannot be invoked to establish specific distinctions. 



I believe that O. ambulator (including 0. carinatus) can not be 

 united with O. fastigatus particularly on account of the dimensions 

 of the radial shields and of the development of the ventral keel, 

 which is never lacking in the first species. 



Matsumoto has just published some figures and notes on the sub- 

 ject of O. fastigatus ('17, p. 328, fig. 91). He records especially va- 

 riations in the form of the adoral plates and the accidental presence 

 of the small accessory mouth shields. He considers Liitken and Mor- 

 tensen's 0. carinatus as a synonym of O. fastigatus, an opinion which 

 I can not share; as I have just explained, O. carinatus is identical 

 with O. ambulator, which is very different from O. fastigatus. 



