OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 347 



pair of pores appeal on all the arms. Everywhere else the pores are 

 lacking. 



In Pectinura elata from the Atlantic, as I stated in my original 

 description in 1907, the ventral pores are always lacking; I give a 

 photograph of the ventral surface of one of the specimens collected 

 by the expeditions of the Travailleur and of the Talisman (fig. 8). 



The tentacle scales are always very large on the first arm seg- 

 ments, and their size decreases rather rapidly from the first pair on- 

 ward, but their dimensions are rather variable; perhaps there is 

 some relation between their size and the presence or absence of the 

 ventral pores. It is not rare to see the scale of the pores of the first 

 pair doubled, and this may even happen on the pores of the second 

 pair. The scales of the first pair may be broadened so as almost to 

 be in contact in the median line of the arms, as I find in two speci- 

 mens from station 5657 (tig. 7) and on that from station 5608 (fig. 5). 

 I have already recorded this considerable development of the first 

 tentacle scale in P. elata collected by the Talisman, and it may be 

 seen in the photograph which I give of it (fig. 8). When these 

 tentacle scales are not in contact, though much broadened, the sec- 

 ond under arm plate tapers rapidly in its proximal region; this is 

 what occurs in the specimen from station 5657, shown in figure 7, 

 and in that from station 5608 (fig. 5). I also find the same feature 

 in the specimen from station 5648, which has a disk diameter of 35 

 mm., as well as in those from stations 5460 and 5613, which are 

 smaller, with the diameter of the disk from 23 mm. to 25 mm. In 

 other larger specimens the tentacle scale is less developed. The more 

 or less large dimensions which this scale may assume are not there- 

 for exclusively correlated with size. 



The form of the mouth shields also shows very great variation. 

 As a general rule these shields are triangular and a little broader 

 than long, their outlines being somewhat modified by the more or 

 less considerable development of the granules which to a varying 

 degree extend over their borders as well as also over the adoral and 

 oral plates. In the Investigator specimen shown in figure 1 these 

 shields are markedly broader than long ; it is the same in the speci- 

 men from station 5494 (fig. 2) and in one of those from station 5495 

 (fig. 11). These shields are almost as long as broad, with some 

 variation possible in the same specimen ; I show here specimens from 

 stations 5114 (fig. 3), 5587 (fig. 6), 5657 (fig. 7), and 5495 (fig. 9) 

 in which the mouth shields are almost as long as broad. These 

 shields may become much elongated and appreciably longer than 

 broad ; this occurs, for instance, in the large specimen from station 

 5608 (fig. 5), in the two from station 5468, and in one of the three 

 from station 5495 (fig. 10). A comparison of the three specimens 

 from this last station is very interesting, for one of them has the 



