66 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



Dr. H. L. Clark found it very easy to prepare perfectly expanded specimens of 

 the comatulids which he found at Maer Island in Torres Straits. They were lifted 

 from the sea water, with the cirri down, and plunged abruptly into strong alcohol 

 in a shaUow, flat dish, care being taken to press the central portion down at once to 

 the bottom of the dish. The contraction of the dorsal ligaments causes the arms to 

 lie out flat against the bottom of the dish. The following contraction of the ventral 

 muscles is occasionally strong enough to bring the arms up over the disk and to 

 get them badly entangled, but in the very great majority of cases it is so feeble and 

 so quickly followed by relaxation that a little manipulation of the arms, smoothing 

 them down with the fingers and pressing out the curves, results in perfectly expanded 

 specimens. 



IDENTIFICATION OF COMATULIDS 



The comatulids are exceptionally difficult animals to identify. Many of them 

 are very variable, so variable that different individuals may present a widely different 

 aspect. This is especially true in those species in which the number of arms 'varies 

 between wide extremes, in which the number of elements in the division series is 

 subject to much variation, and in which the arms vary very greatly in relative length. 

 It is particularly true in the comasterids. 



It not infrequently happens that two comasterids with the same number of arms 

 belonging to species of different genera will resemble each other more closely than 

 either resembles others of its own species with a different number of arms. 



In those groups in which the majority of the species have numerous arms and 

 distinctive division series the species with 10 arms are often confusing. This is 

 especially true in the comasterids, in which we are accustomed to give great weight 

 to the characters offered by the division series and in which, therefore, the differences 

 in the cirri, pinnules, and other structures, though constant, seem at first sight to be 

 trivial and unimportant. 



The difficulties attending the determination of the comatulids are greatly 

 increased by the fact that practically all specimens examined are more or less exten- 

 sively mutilated. 



In identifying comatulids, therefore, it is necessary to follow the keys very closely 

 and literally. The short cuts, which can be used successfully in the case of nearly all 

 other creatures, are likely to lead one astray in the case of the comatulids. The 

 general appearance especially is deceptive. 



TREATMENT OF THE SPECIES 



At the head of the account of each species is an annotated list of references 

 including all the works which the author has consulted. These references cover 

 not only the synonymy and distribution, but also the historical, anatomical, and 

 other aspects of the study of these animals which were included in Parts 1 and 2 

 of Volume 1. All of the references have been personally verified. 



Following the references is a brief summary of the essential characters which 

 serve to distinguish the species from closely related types, and sometimes also from 

 types not closely related but superficially very similar. 



