PART 5 A MONOGRAPH OF THE EXISTING CRENOIDS 825 



to each radial area, the pairs of columns in each radial area being separated from each 

 other proximally by more or less marked interradial ridges. In the young, each cirrus 

 socket is bordered by a strongly developed ventrolateral horseshoe-shaped ridge rather 

 higher at its two ends than elsewhere, the two ends being somewhat projecting. In 

 older animals this ridge decreases in height pro-ximally (ventrally), where it finally 

 becomes obsolete, witli the result that the cirrus sockets are merely flanked on each 

 side by a prominent lateral tubercle longitudinally elongated and triangular in side 

 view. These tubercles give the appearance of interrupted longitudinal ridges bordering 

 each column of cutus sockets. Though well separated in the j'oung, the cirrus sockets 

 later gradually become larger with the result that in the largest examples thej^ are 

 closely crowded. As the sockets of one column alternate in position with the sockets in 

 the columns on either side, the lateral tubercular prominences of neighboring columns 

 finally come to lie in a straight line and the cirrus sockets are separated all around the 

 centrodorsal by prominent serrate ridges, their proximal and distal borders not being 

 marked by prominences or ridges of any kind. 



The cirri are very rarely recovered. A cirrus which appears to lack but a single 

 distal segment measures 30 mm. in length and is composed of 28 segments (Carpenter 

 records about 35 in a cirrus, presumably from a Blake specimen since the Challenger 

 specimen had none), of which the first four are very short and the following are four or 

 five times as long as their proximal widths, gradually becoming shorter distally with the 

 result that the last two are scarcely tmce as long as broad. The segments all have 

 prominent ventral ends which more or less overlap, this feature being most pronounced 

 in the distal half but becoming less marked again terminally. Excepting for the proxi- 

 mal segments, the cirri are very strongly compressed laterally. The dorsoventral 

 height of the segments increases somewhat in the outer half, gradually decreasing again 

 in the terminal portion with the result that the last 6 segments taper rapidly to a point. 

 According to Carpenter, the terminal claw is very small. In a specimen which was 

 taken by the Atlantis off Cuba and which H. L. Clark attributed to Atelecrinus pour- 

 ialesi, the bases of some cirri were left intact as far as the sixth segment. Dr. Clark 



Figure 53. — Atelecrinus balamides P. H. Carpenter, paratype, B.M., 

 88.11.9.1, ChalUnger station 122. 



4 mm 



