348 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



mouth shields almost as long as broad (fig. 9), the other has them 

 a little longer than broad (fig. 10), and in the third they are broader 

 than long (fig. 11). I have shown these three specimens side by 

 side (figs. 9-11). In P. elata from the Talisman collection the 

 mouth shields are as long as broad (fig. 8). 



As for the supplementary plate which may be found beyond the 

 mouth shield, it appears in certain individuals and is lacking in 

 others ; it may even occur in two or three interradii and be lacking 

 in the others in the same individual ; there does not seem to be the 

 least relation between the presence of this plate and the size of the 

 specimen. 



If we examine the arm spines, we find that it is only in the small 

 specimen from station 5114 that they do not exceed three at the base 

 of the arms (fig. 15), and this number falls to two at a little distance 

 from the disk ; but on the two or three first arm segments sometimes 

 I find a very small fourth supplementary spine, which appears at 

 the superior angle of the side arm plate. In the other specimens in 

 which the diameter of the disk varies between 19 mm. and 25 mm. 

 there are always some arm segments at the base of the arms which 

 have four spines, and in the larger specimens this number reaches 

 five, falling afterwards to four and in the distal half of the arms 

 to three. In the small specimen from station 5114 the three spines 

 are equal; in that from station 5587 (fig. 4) the four spines are 

 short and equal. In the other specimens the spines are generally a 

 little longer, and the ventral spine slightly exceeds the others. Some- 

 times the arm spines are rather long at the base of the arm, and 

 they may even reach the length of the segment ; I find this condition 

 in the specimen from station 5505 (fig. 13), but usually these spines 

 do not exceed half the segment, and they become smaller as the dis- 

 tance from the disk increases. In the large specimens the number five 

 is frequent at the base of the arms and is maintained for a variable 

 length, it falling to four and finally to three, as I have just said, but 

 with rather frequent irregularities. Here again the variations are 

 very great. I may recall that the presence of five arm spines, with 

 the ventral the largest, was one of the principal characters upon 

 which I based my Pectinura modesta. In Pectinura elata the spines 

 are four in number at the base of the arms, and then three for a 

 rather large portion of their length ; but I sometimes find five at the 

 base of the arms, though that is very rare ; these spines are subequal, 

 but the ventral spine is sometimes a little larger than the others. 



Considering all these variations, I believe that B. conspicua should 

 be considered as a rather polymorphic species, and I no longer be- 

 lieve myself justified in maintaining the two species which I estab- 

 lished on characters which to-day appear inadequate — P. modesta, 



