586 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



the right posterior radial, until its left border comes to lie over the median apex of 

 the posterior basal, and therefore near the right lateral angle of the left posterior 

 radial. 



From this point the course of the radianal is almost directly upward, and as it 

 passes upward the proximal portion of the right posterior radial rapidly extends 

 itself along the edge of the posterior basal beneath it, finallj^ reaching the right 

 lateral angle of the left posterior radial and entirely excluding the radianal from 

 contact with the posterior basal. 



When the posterior radials have become completely joined so that the posterior 

 interradial section of the radial circlet does not differ from the other interradial 

 sections, the radianal is seen to lie upon the surface of the ventral perisomic disk 

 at the base of the anal tube, with the greater part of its area to the right of the pos- 

 terior interradial suture, and with its posterior (proximal) border usually just in 

 contact with the distal border of the radial circlet. 



The raidianal may be resorbed in this position, or it may move slightly farther 

 outward. 



No matter at what stage the radianal is first formed it always follows the same 

 course, and in those cases in which it appears when the radials are almost in con- 

 tact, or even after the complete closure of the radial circlet, its behavior is the same 

 as if it had been present before the appearance of the right posterior radial. 



The radianal is of much slower growth than the other plates of the crown, and 

 is always distinguished from them by a much finer structure. 



Interradials. — The presence of interradials has been determined in the penta- 

 crinoids of four species of comatulids, in which they occur in two very different 

 forms. 



Interradials are formed on the distal border of the radial circlet in Comac- 

 tmia meridionalii, Comissia littoralis, and in Antedon bifida. 



In Comactinia mei'idionalis and in Gojiiissia littoralis they first appear when 

 the radials have almost come into contact, at about the time of the formation of 

 the IBr^, as small plates in the triangular space above the approximated lateral 

 angles of the radials and below the proximal borders of the orals. They grow 

 rapidly into conspicuous rhombic plates the distal borders of which are on a level 

 with the bases of the IBrj. 



In Comactinia, mendionalis, as the space between the proximal border of the 

 orals and the distal border of the radial circlet increases additional interradials 

 appear, so that there may be as many as five or six, which grade into the perisomic 

 plates of the disk. 



The further development of the interradials in Comissia littoralis is not known. 



In many of the Comasteridse (for instance, in Co77ianthina schlegelii and in 

 certain large species of Comaster) the dorsal interradial areas as far as the edge of 

 the disk are occupied by a solid calcareous plating, which often is of considerable 

 thickness. It is possible that this is a development from the larval interradials 

 which have been retained. 



In Antedon bifida the interradials are very rudimentary and not always 

 present; they appear at about the time of the formation of the IBrj and appar- 



