646 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



Hartlaub recorded a specimen from Fiji in the Hamburg Museum having eight 

 of the IIBr series 2 and two 4 (3+4), these last being on a IBr 6 (5 + 6) series. Evi- 

 dently this is one of the specimens which I found labeled "Moreton Bay." Hartlaub 

 gave only 2 from that locality, whereas I found 3. It is possible, of course, that he 

 considered Moreton Bay as in Fiji, following Carpenter. 



Muller wrote (1849) that the specimen from Vavao, Tonga Islands, apparently 

 belongs to parvicirra. The centrodorsal is very small in comparison with the width 

 of the division series, since not only the radials but also the IBr, are appressed against 

 the corresponding ossicles on either side. Likewise the first brachials are in contact 

 exteriorly with their neighbors, the arms being thereby increased in width. 



The centrodorsal is flattened, and the slender marginal cirri are few in number. 



The radials are visible. The IIBr and IIIBr series are 4 (3 + 4). 



There are 35 arms. The brachials are smooth, with everted distal edges. 



The pinnules of the division series are long. P 2 or one of the following pinnules 

 is very small; then the length increases. 



The disk is smooth. The color in alcohol is yellow. 



Carpenter (1879) said that he found this specimen in the Paris Museum under 

 the manuscript name of Comatula brevicirra Troschel, and that it differs from poly- 

 morpha (timorensis + parvicirra) in many subordinate characters, and is not absolutely 

 identical either with the type or with any one of the 4 varietal specimens which he 

 had examined. He said that he could not but regard it as representing another of 

 the slight and probably very numerous modifications of this type (polymorpha) , 

 of which he believed it most likely that Muller's original species (parvicirra) is also 

 a varietal form. 



In 1881 Carpenter said that both in this specimen and in the small ones obtained 

 by Pe'ron and Lesueur (in the Australian Seas) the brachials overlap much more 

 strongly than they do in Philippine examples, the distal edges of successive brachials 

 being much more raised. 



I examined this specimen in 1910. The centrodorsal is very irregular in shape. 

 The cirrus sockets are partially obliterated. There remain only small basal portions 

 of 4 cirri. 



There are 35 arms. Three of the postradial series divide three times, thus bearing 

 8 arms, and the other 2 are in process of regeneration. One of the latter, the anterior, 

 has 2 + 3 arms, with the IIBr series 2 and the single IIIBr series 4 (3 + 4); the other, 

 the left posterior, has 2 + 4 arms, with the IIBr series 2, and the IIIBr series, both of 

 which are on the same IIBr series, 4 (3 + 4). The division series are broad, and the 

 arms are short and broad. The brachials have strongly produced distal ends. 



This specimen probably should be referred to timorensis rather than to parvicirra. 



The Challenger specimen from the Tongatabu Reefs was described by Carpenter 

 as a new species under the name of Actinometra quadrata. He gives the characters 

 of this new form as follows : 



The centrodorsal is a small thin disk bearing a single row of marginal cirri. 



The cirri are X, 11; the fourth and fifth segments are slightly the longest. 



The radials are largely visible. The IBr! are partly united laterally, but the 

 division series are quite free. The IIBr series are 4 (3 + 4). 



