ME. P. H. CAEPENTER OX THE GENUS ACTINOMETEA. 43 



arc disposed some 20 or 25 cirrlii iu one row, but with occasional traces of a second, 

 in which the civrhi alternate in jiositiou Avith those of the first row. The number of 

 segments in each cirrhus is normally from 11 to 14, of which the last forms a recurved 

 claw, while a more or less distinct spine is usually visible upon the dorsal edge of each of 

 the three or four penultimate segments (PI. III. figs. 8-11). 



In PI. II. fig. 8, is seen an abnormal condition of the centrodorsal piece, which is of 

 an irregular oval form, and so extended as to conceal large portions even of the second 

 radials. These last are usually more or less completely united with one another late- 

 rally. The amount of their union is to a certain extent dependent uj)on the number 

 of arms developed. Thus in the small specimen with only 13 arms, represented in 

 PI. II. fig. 9, the second radials are not united laterally for more than half their length ; 

 in fig. 10 (2G arms) the union is somewhat more complete, and even more so in fig. 8 

 (28 arms), while in the variety with 39 arms, represented in fig. 11, the second radials 

 are completely and closely united with one another all round. This rule, however, 

 appears to be only a specific one, and not generally applicable to all Comatulcs ; for in 

 the 80-armed JPhanogenia the second radials, as figured by Loven ', do not appear to 

 be united with one another any more closely than they are in the small 13-armed speci- 

 men of Act. 'polymorpha (PL II. fig. 9). 



In Act. potymorpha the two segments (first distichals, palmars, or brachials) borne by 

 any axillary are united to one another laterally to about very much the same extent as 

 the second radials are ; i. e. when the number of arms is small, their first segments, 

 whether primary, secondary, or tertiary, are not laterally united in pairs with such com- 

 pleteness as when the division of the ten primary arms is carried to any considerable 

 extent (PI. II. figs. 8-11, d„ p„ ftj. 



When the arm-division is unequal it is generally carried further in the trivium or 

 posterior radii, C, D, E, than in the two anterior radii. A, B, which form the bivium. 

 This is well seen in PI. II. fig. 9, in which no distichals are developed on either of the 

 two anterior radii ; and again in fig. 10, in which, while distichals are developed all 

 round, the division is carried no further in one of the anterior radii, while in each of the 

 others from one to three palmar series may be developed. In only four normal cases 

 have I found an anterior radius to bear more arms than a posterior one. In each of 

 these the total number of arms was considerable, and one at least of the two posterior 

 radii bore the same number of arms as the abnormal anterior one. Thus, for example, in 

 PI. II. fig. 11, each of the radii bears eight arms, with the exception of the posterior one 

 (D), on which only seven are developed. This, however, is an abnormal case of fracture 

 of the whole radius between its second and third segments. The new portion is con- 

 siderably smaller than the old, the proximal articular face of the new axillary l)cing far 

 less wide than the corresponding distal face of the old second radial ; while both the 

 distichal series which it bears are imperfect and abnormal, so that the absence of a 

 further division in one of the secondary arms is not particularly remarkable. 



(§ 30) The number of arms that may be developed in Act. polyniorpha is a character 



' " Phanofjenia," loc. cit. p. 230. 

 SECOND SERIES. — ZOOLOGY, VOL. II. 7 



