684 PROCEEDINGS OF THE NATIONAL MUSEUM. voi. xxxm. 



Doctor Carpenter, in his report on the " Comatulse " of the (7Aa?- 

 ?<;'y;-(/e'/' expedition, divided Comatula {=Actinometra) as follows: 



Series 1 ; the two outer radials and the first two brachials united by syzygy : 



Ten arms Solaris group. 



Two distichals united by syzygy Paiicicirra group. 



Three distichals, the axillary a syzygy Typlca group. 



Series II; the two outer radials articulated: Ten arms Echinoptera group. 



Series III; two articulated distichals: 



Palmars and post-palmars like distichals; but first two brachials united 

 by syzygy SteUigcra group. 



First arm-syzj^gy in the third brachial VnUda group. 



Series IV; three distichals, the first two articulated, and the third axillary 

 witli a syzygy. 



First arm-syzygy in the second brachial , Fimbriata group. 



First arm-syzygy in the third brachial Parvicirra group. 



This arrangement of the species was merely intended as a con- 

 venient guide to their identification, and nothing more was claimed 

 for it. Its artificial character may be judged from a single species, 

 the Alecto pariricirra of Johannes Miiller 1841 {^^Actmometra parvi- 

 cirra of Carpenter) , the type species of Carpenter's last group. This 

 species is not infrequently ten-armed, therefore falling in the ^^Echi- 

 noptera group ;" again, it may have all the distichal series of two 

 articulated segments, the first arm-syzygy falling in the " third 

 brachial," in which case it belongs with the ^'"Valida group" according 

 to Carpenter's scheme; yet the species is made the type of a third 

 group, the ^''Parvicirra group." The groups themselves, contrary 

 to what was the case in the various " groups " and " series " of ^^Anfe- 

 (1,071,'''' are, with a single curious exception, the ^'Typica group," fairly 

 homogeneous. Carpenter gives four species as belonging to this 

 group, distincta, typica, novm-guincn', and multibrachiata, and a 

 fifth, (/racilis, has since been described. In distincta and inidti- 

 hrachiata, the costals are united by syzygy, according to the descrip- 

 tions of these species, the distichals are 4 (3-f4), and the palmars 

 2 (1-1-2) ; but on Plate lv and Plate lvi quite a different arrange- 

 ment is fomid; while the palmars are 2 (l-|-2), the distichals are 

 4 (l-)-2; 3-|-4) ; in other words the distichals, instead of being 

 " three, the axillary a syzygy " are four, united in tioo syzygial pairs. 

 This simplifies matters considerably, as will be seen further on. 

 Now, in typica and gracilis, and in novce-guineoi as well as I can 

 judge from the figure, the costals and first two distichals are not 

 united by sj^zygy; while the union is very close, it is of the same 

 type as that between the costals in such species as Carpenter's 

 Acti7iometra rohustipinna and Miiller's Alecto parvicirra and Alecto 

 trichoptera. We find, then, that three of the five species of the ^^ Typica 

 group," including typica itself, fall in " Series IV," while the other 

 two do not belong in Carpenter's scheme at all. 



