308 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



longer than broad, with the proximal angle strongly marked and 

 sharp. 



I shall record also the variations shown by the tentacle scales. 

 These are typically two in number, the outer rapidly becoming 

 very much larger than the inner. These two scales may remain 

 subequal on the first arm segments and the internal scale may not 

 become really reduced until beyond about the tenth; this is what 

 occurs in the specimens from stations 5582 (fig. 3), 5585 (fig. 7), 

 and 5650 (fig. 6), but in that from station 5348 (fig. 8) this in- 

 ternal scale becomes very much smaller than the other beyond the 

 second segment, and it is considerably reduced on the following 

 segments. On the other hand, this same scale may be doubled; 

 this occurs over a large part of the length of the arms in the speci- 

 men from station 5582, which possesses three tentacle scales (fig. 

 11), one very large and the two others very small. 



The adoral plates are much elongated and more or less narrow, 

 but I do not find any variations so great as those which have been 

 described by H. L. Clark, and which may go as far as the com- 

 plete disappearance of these plates through a sort of reciprocity 

 with the oral plates. 



The determination of the rather important variations which I have 

 just noted in O. fastigatus has led me to take up again the study 

 of the comparative characters of this species and of O. ambulator ^ 

 which I established in 1896 from specimens dredged by the Investiga- 

 tor in different localities in the Indian Ocean, at depths varying be- 

 tweeen 200 and 890 fathoms. Before inquiring into whether 0. am- 

 bulator does not represent simply one or other of the variations 

 which I have just described in O. fastigatus, I may say that I con- 

 sider O. carinatus, described by Liitken and Mortensen three years 

 after my description of O. ambulator, as being absolutely identical 

 with it. I had already foreseen this in 1904 ('04, p. TO). 



One of the principal characters which distinguishes O. ambulator 

 from O. fastigatus is the large size of the radial shields, for these 

 shields are only a little shorter than half the radius of the disk ; this 

 is the same proportion which is shown by Liitken and Mortensen in 

 their 0. carinaftus. Whatever may be the variations shown by the 

 radial shields in 0. fastigatus, they always remain very reduced in size 

 and never reach the dimensions which they show in 0. ambulator (fig. 

 12). Other characters also show that O. ambulator is a species very 

 distinct from 0. fastigatus. The keel which extends throughout the 

 whole length of the ventral median line of the arms in O. ambulator 

 occurs with very great constancy ; it shows a very remarkable devel- 

 opment and is very much more accentuated than in- O. fastigatus, in 

 which it may sometimes entirely disappear. The character furnished 



