L'L'S nil. 1. 1 . ll\ 82, I Ml I Kl> STATES NATIONAL MUSEl'M 



are exceedingly variable, generally or locally, in all their structural characters and 



in their size. 



History.— It is possible thai the Alecto horrida described by Dr. W. E. Leach 

 in 1 S 1 "> . which i- i In' typo species of die genus Alecto, represents the species described 

 by Prof. Johannes Muller in 1846 as Comatula (Alecto) reynaudi. But Leach's de- 

 scription U very vague, his figure is not with certainty identifiable, and his type 

 specimen ran not be found, so that the identity of horrida with reynaudi rests on 

 quite inconclusive circumstantial evidence. 



The final indubitable species of this genus to be recorded was assigned by Savigny 

 io the genus ( 'omatvla; it was figured in his account of the zoology of Egypt published 

 in L826 under the editorship of Audouin. In 1841 Prof. Johannes Muller referred 

 [lie same species to Alecto, and in 1S4G ho described a second species under the generic 



nation Comatula (Alecto). 



In 1879 Dr. P. H. Carpenter assigned savignii, philiberti, and reynaudi to the 

 genus Antt dan as understood by him. In 1882 he referred to the species of this type 

 as belonging to the Savignyii group. In 1888 he defined the Savignyi group and 

 included in it reynaudi, savignyi, and philiberti, the recently described variipinna, 

 aeutieirra, ludovici, and bipartipinna (the last three falling in the genus Craspedo- 

 metra as now understood), and the new species anceps, quinduplicava, and angustiradia 

 (the last a species of Adelometra). As one of the three specimens had only 10 arms, 

 Carpenter also included Antedon anceps in his Milberti group. Antedon clemens he 

 included in his Palmata group which consisted of species with the IIBr series 2. 



Hartlaub in 1891 followed Carpenter in placing clemens in the Palmata group, 

 and included in the Savignyi group angustiradia, reynaudi, savignyi, anceps, variipinna, 

 quinduplicava, ludovici, bipartipinna, and philiberti, and his new species bengalensis, 

 martensi, kraepelini, brockii, affinis, nematodon, and crassipinna. 



The Savignyi group of Carpenter and Hartlaub was therefore the equivalent of 

 the very closely allied genera Heterometra, Himerometra, and Craspedometra combined. 



In my first revision of the old genus Antedon, which was published in 1907, the 

 species of Hi t< ronu tra were all referred to the new genus Himerometra. 



In a revision of the family Himerometridae published in 1909, the genus Himero- 

 metra was redefined in such a way as to include only those species in which P D is larger 

 and longer than P l7 which in turn is larger and longer than P 2 . The species with 

 1 1 Br 4(3+4) series and therefore more than 10 arms, and P D smaller than P 1( were 

 distributed among three genera. The species in which the middle and distal brachials 

 are exceedingly short and discoidal and the elements of the division series and lower 

 brachials are swollen were assigned, together with a number of 10-armed species, to 

 the new genus Amphimetra. These species were anceps (= quinduplicava), producta, 

 Bchlegdii, and rariipinna (in the sense used by Carpenter, that is, including both 

 variipinna and crenulata). The species with the middle and distal brachials not 

 especially short, more or less obliquely wedge-shaped, and the elements of the division 

 series and earlier brachials not swollen, were assigned to the new genus Heterometra. 

 These species wore affinis, bengalensis, brockii ( = variipinna) , quinduplicava, reynaudi, 

 and savignii. The species with the cirri tapering distally, the distal cirrus segments 

 twice as long as broad, and no opposing spine, were assigned to the new genus Cras- 

 pedometra. The only species at that time included in Craspedometra which is of present 



