322 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



at the tip of the axillary from which they arise, while those on the outer side broaden 

 out and cover the IBr series except for a narrow median area. Even the centrodorsal 

 may be purplish. The disk and oral surface of the arms are yellowish, quite yellow 

 when dry. 



Notes. — A narrow and low carLnation runs along the arms to the tip in all the 

 specimens which I have examined. 



The example from Port Molle is a fine representative of the species. The centro- 

 dorsal is reduced to a pentagonal plate which has not quite sunken to the level of the 

 radials. 



Localities.— Alert; Port Molle, Queensland, 22 meters [Bell, 1884; P. H. Car- 

 penter, 1888; A. H. Clark, 1913; H. L. Clark, 1916] (1, B. M.).' 



Endeavour; 8 miles east of Sandon Bluffs, New South Wales; 64-73 meters 

 [H. L. Clark, 1916] (4, M. C. Z., 709, 721). 



Remarks. — In his discussion of Comat'ula Solaris in the Challenger report (1888) 

 Carpenter remarked that the centrodorsal of the large Vienna specimen (the type of 

 Actinometra imperialis J. Miiller, 1841) has lost all trace of the cirrus sockets on one 

 side and is almost reduced to a level with the radials, while in an Alert specimen from 

 Port Molle the cirrus sockets are all obliterated, leaving nothing but a thin flat plate, 

 very much as in some forms of Comatiila rotalaria. He noted, futher, that in this 

 specimen from Port Molle the disk is perfectly soft and membranous. 



I examined the specimen from Port Molle in 1910 and published a note on it 

 in 1913. 



In 1916 Dr. Hubert Lyman Clark in his description of Comatula cratera remarked 

 that whether this fine comatulid should be considered distinct from C. Solaris is, of 

 course, a matter of opinion. He recalled that the locality where it was dredged is 

 some 400 miles south of the most southerly locality from which C. Solaris has been 

 recorded, and that the depth is considerably greater than any depth which has been 

 published for that species. 



In view of these facts he believed that the absence of cirri and the reduction of 

 the centrodorsal, combined with the difi'crent terminal combs of the oral pinnules 

 and the absence of plating on the disk and on the oral surface of the arm bases, 

 warrant the recognition of the form by a different name. 



In C. Solaris the terminal comb of Pj, according to Doctor Clark, consists of 

 35-40 segments and occupies nearly half the pinnule, while the shape of the individual 

 teeth is, moreover, very different from that seen in the combs of C. cratera. 



He says that the specimen dredged by the Alert at Port Molle and referred to C. 

 Solaris by Carpenter, but specially discussed because of the absence of cirri and the 

 unplated disk, is very possibly a specimen of C. cratera; but he does not mention the 

 suppression of the cirri in one-half of the centrodorsal in the type of Actinometra 

 imperialis. 



In 1918 I regarded the specimens both with and without cirri as representing C. 

 Solaris, and so assigned them in the key to the species of Comatula published in the 

 Siboga report. I had not at that time seen Doctor Clark's description of C. cratera. 



