m 



ore slender, but otherwise similar. P 3 is similar to P 3 , but slightly smaller and shorter. P is 

 3.5 mm. long with 17 segments of which the distal bear traces of a comb. P. is very small 

 and slender, 3 mm. long, with 110 tracé of a comb. 



In the " Challenger" report Dr. P. H. Carpenter records (p. 304) under the name of 

 Actinometra pulchella a somewhat anomalous specimen which was dredged near the Kei Islands 

 (5 49' 15" S., 1 32 14' 15" E.) in 140 fathoms. Actinometra pulchella (now known as Neo 

 comatella alata) is confined to the Caribbean region, and belongs to a genus, Neocomatella, 

 which is exclusively Atlantic, its representative in the Indo-Pacific region being the allied 

 Comatella. 



This specimen gave Carpenter considerable trouble ; at first he had believed that it 

 represented a new species, and on p. 93 of the "Challenger" report he mentions it under 'the 

 name of Actinometra diffJcilis, saying of it that "the two outer radials [i. e., the two elements 

 of the IBr series], the two distichals [i. e., the two elements of the IIBr series] and the first 

 two brachials are respectively united by syzygy." He refers to fig. 2 on pi. 52 as representing 

 this specimen. 



The figure shows a much broken individual with about 1 8 arms ; the dorsal pole of 

 the centrodorsal is circular, broad and fiat, 4.0 mm. in diameter ; there were apparently about 

 XX cirri which are 13 mm. long with 16 segments of which the longest are over twice as 

 long as broad, and the last eight are broader than long ; the brachials beyond the basal 

 appear to be triangular, about as long as broad, judging from the regenerating arm in the 

 upper right hand side of the figure. 



It is evident that Carpenter's specimen is very close to that just described, differing 

 chiefly in the larger size, and in certain features such as an increased number of cirrus seg- 

 ments and a relatively lesser length of the longer proximal cirrus segments, which are usually 

 correlated with increased size. The apparent difference in the number of the cirrus sockets maj- 

 or may not be real, on account of the indistinctness of the drawing. It seems reasonable, 

 therefore to refer the "Siboga" specimen to the same species for which the name diffïcilis is 

 available. 



With the arm structure of the Atlantic type [Neocomatella] this species possesses a 

 very anomalous centrodorsal which is circular, and has the second row of cirrus sockets below 

 the first instead of alternating with it as in all of the other genera of the Comasteridac. 



It appears most logical, therefore, to erect for the reception of this species a new genus, 

 Palaeocomatella, most nearly related to the Atlantic Neocomatella, but differing markedly in 

 its curious centrodorsal. Whether this disposition is the correct one must be left for future 

 investioration to determine. 



*> 



Capillaster A. H. Clark. 



Key to the Species of the Genus Capillaster. 



a 1 Cirri absent ; centrodorsal reduced to a small stellate plate lying within the radial circlet 

 (Borneo and the China Se a) macrobrachius 



SIBOGA-EXPEDITIE XLI1<5. 2 



