494 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxxvi. 



The history of the family Comasteridte may be said to date from 

 that most excellent memoir of Dr. P. H. Carpenter on the genus 

 Actinometra. In this memoir he gives a detailed account of the 

 systematic treatment of the various species of the genus by previous 

 authors, and ably reduces to order a systematic chaos scarcely sur- 

 passed in the whole subject of zoology ; for even so great a zoologist 

 as Prof. Johannes Midler, in the only monograph then published on 

 the unstalked crinoids, had ]3laced a single species (under four dif- 

 ferent specific names) under the subgenus Actinometra among the 

 exocyclic forms (twice) ; under the subgenus Alecto among the endo- 

 cyclic forms, and again under the subgenerically incertce sedis, in 

 the heterogeneous group Comatula. 



Since 1879 the genus Actinometra has been accepted in the sense 

 in which it Avas used by Carjjenter. Nine years afterwards he split it 

 up into eight specific groups, distributing these among four " series," 

 and this arrangement has been used ever since. Shortly afterwards 

 the genus was raised to family rank, on a par with the family "An- 

 tedonidse,"' covering the genus '"''Antedon " of Doctor Car[3enter. 

 This I have recently shown to be an unnatural division. In the 

 course of my work I found the genus Actinometra becoming some- 

 what unwieldy, and I accordingly split it in two sections, one small 

 and one large; but with the accession of new material the large divi- 

 sion proved not to be natural, and I split that into two parts. 

 Enormous collections from the Philippine Islands having been re- 

 ceived, a still further change was seen to be necessary, and the last 

 of the divisions created was shattered into three fragments. I dis- 

 carded the appropriate and euphonious name Actinometra proposed 

 b}^ Professor Miiller in favor of Comatida of Lamarck, of Avhich it is 

 a pure synonym, with the same type. While the name Comatulida^ 

 was first employed to cover the family in place of Actinometridsc 

 (not available because of the disuse of Actinometra) , I soon found 

 that confusion with the Comatuladae of Fleming (1828) and Comatu- 

 lidse of d'Orbigny (1852) and succeeding authors, with a more or less 

 comprehensive range of meaning, but never so restricted as to cover 

 the "Actinometrida^ " alone, made a change desirable, and I there- 

 fore substituted " Comasteridse," the name being derived from that 

 of the next oldest genus. 



The most important discovery made in regard to the Comasteridse 

 since the publication of Carpenter's memoir in 1879 is that of Mr. 

 Frank Spi'ingei-, who in 1903 described and figured a strongly de- 

 veloped ambulacral plating on the arms and pinnules of a new species 

 from the Tortugas. I have since found these plates to be universally 

 present in the species of the " Fimbriata group " from the West 

 Indies, well developed even in '■'■ Actinometra " lincata, in which I de- 

 tected it in some of the Challenger specimens previously examined by 



