346 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



Crinometra pulchra, C. margaritacea, C. concinna, C. insculpta, and C. gemmata. In a 

 paper published on November 16, 1909, I wrote: 



Carpenter was loth to believe that a single species could be both "bi-" and "tri-distichate" ; he 

 observed the more or less frequent occurrence of both forms in the same individual, but he assigned the 

 specimens to their systematic position on the strength of the character of the majority of their series. 

 This is usually a safe enough guide, but in a few cases, such as this [Comanthus parvicirra ("rotalaria")], 

 it is unreliable, and leads to erroneous conclusions. Minckert, on the other hand, had gone too far 

 in the opposite direction ; he credits a certain Caribbean species of Charitometrinae, Crinometra brevipin- 

 na, with the possession of IIBr series which may be all 4(3 + 4), all 2, or both in any combination, and 

 he even makes it the type of a new "group", the "Brevipinna group". In reality he has confused a 

 number of perfectly distinct species some of which have one arrangement, some the other, none, 

 however, having both in equal proportions. Crinometra in this respect is exactly like Pachylometra [in 

 the broad sense], its representative in the East Indies. 



The extensive collections brought together by the United States Coast Survey 

 steamers, particularly the Blake, together with a few from other sources, as for instance 

 those that had been taken by Capt. A. Cole of the cable repair ships Enterprise and 

 Investigator, had been sent by Alexander Agassiz to Dr. P. H. Carpenter for study and 

 report. Some time after Dr. Carpenter's death on October 21, 1891, they were reas- 

 signed to Dr. Clemens Hartlaub. One of Hartlaub's students, Dr. Wilhelm Minckert, 

 used the material in connection with a paper published in 1905, but the complete report 

 was not published by Hartlaub until 1912. 



Dr. Hartlaub was a friend and admirer of Dr. Carpenter, and his work on the Blake 

 comatulids was modeled after the Challenger report. He told me that he considered it 

 his duty to do this as the work had originally been undertaken by Carpenter, and also 

 he wished to make his report conformable to the treatment in the Challenger report. 



His account of this species is most excellent and exhaustive. He could not quite 

 bring himself to adopt Minckert's concept of a Brevipinna group including species in 

 which the IIBr series may be either 4(3+4) or 2, or both in any combination, but re- 

 verted to Carpenter's classification in which those species with plated ambulacra 

 sharply flattened sides, and all or a majority of the IIBr series 4(3+4) were assigned to 

 the Granulifera group and those, otherwise similar but with the IIBr series 2, were 

 assigned to the Spinijera group. 



He was thus led to distribute his specimens among three species, angusticalyx and 

 granulifera in the Granulifera group, and brevipinna in the Spinijera group; but he dis- 

 cussed some specimens that seemed to him to be intermediate between granulifera and 

 brevipinna, although he could not bring himself to admit the specific identity of the two 

 forms. In identifying certain specimens with the Indo-Pacific angusticalyx, a species 

 known to him only by Carpenter's description and figures, he was undoubtedly influenced 

 by Carpenter's extensive comparisons between angusticalyx and granulifera in the Chal- 

 lenger report. 



He divided brevipinna into twelve varieties. Of the twelve new names proposed by 

 him seven (gracilis, pulchra, elegans, tuberosa, diadema, laevis, and spinosa) had previous- 

 ly been used in the genus Antedon. But since, priority in nomenclautre is not concerned 

 with names of a lower category than subspecies, his varietal names are used herein with 

 the exception of pulchra, which conflicts with Crinometra pulchra, 1909, representing a 

 different variety. 



