48 



' TIIKTIS" SCIENTIFIC KKSULTS. 



Caipenter placed the species typica, diatincta. mul librae! data 

 and liovcegiunecg in his " Typica gi'oup " coinposed of species ia 

 which the first two post radial ossicles (" two outer radials ") are 

 supposedly united by syzygy instead of by^ the usual synarthry, 

 and the same is true of the divission series, which are typically 

 coinposed of two segments each. Now the occurrence of a true 

 syzygy in these places is theoretically impossible, and I therefore 

 seized the first opportunity of examining into the matter. I find 

 that the union between these ossicles is in reality a modification 

 of the primitive synarthi-y in the direction of the syzygy, though 

 tlie dorso-ventral bar across the joint face is never entirely 

 obliterated. All grades of perfection of tliis " pseudo-syzygy " 

 may he traced, from the perfect synarthry of Comaster viultifida 

 to the nearly perfect syzygy of 6*. typica. Tlie " Ty-pica group," 

 then, is shown to 1)6 non-existent as a natural unit. But the 

 memhers of the " Typica group " all agree in their jieculiar combs 

 on the proximal pinnules ; these combs are short with very long 

 teeth, and are abruptly bent so that they do not lie in the geneial 

 line of the pinnule ; moreover, the combs are not confined to the 

 proximal pinnules, but occur at intervals all along the arm to 

 tiie tip, a condition quite unknown in other Comasterids. Also, 

 the proximal pinnules aie very slender instead of being long and 

 stout. In these characters IMiiller's Alecto multifida agrees with 

 the species in Carpenter's "T3fpica group," and must therefore be 

 placed with them, together witli several curious species recently 

 described by Dr. Hartlaub and by myself ; all these forms I have 

 united under the generic name Comaster. 



The determination of variabilis as a Comaster at once brought 

 out the fact that there is absolutely nothing in Bell's description 

 or figure whereby this species may be differentiated from the 

 Actlnometra multibrachiata described four years later by Car- 

 penter, of which I have numerous specimens from the Philippine 

 Islands. The chief character of the latter is the large and 

 rolmst cirri, but Bell says nothing whatever about the size of the 

 -cirri of variabilis, and they are not shown in his figure. 



I requested Professor Bell to compare the types of these two 

 species, both of wliich are in the British Museum, and to inform 

 me whether or not my suspicions as to their identity were well 

 founded. He very courteously replied tliat the figures given of 

 them appear to him " to be quite satisfactoiy and to show the 

 complete diff^erence between them. In A. multibrachiata the arms 

 are free and the cirri are stout and well pronounced, while in A. 

 variabilis the arms are united for a considerable distance and the 

 cirri are either completely absent or very delicate." The figure 

 of variabilis, however, shows the arms to be " free," exactly as 

 in multibrachiata, and, in fact, in all the species of the genua. 



