60 ]MK. P. H. CAllPENTER ON THE GENUS ACTINOMETEA. 



which is more distinctly flattened than in the type, although the small opposing spines 

 ou the dorsal margins do not appear at all until the three or four penultimate segments; 

 and eveu on these tbey are but sliglitly developed. 



In Yar. 2 there are only ten cirrlms-sockets around the margin of the pentagonal 

 centrodorsal plate, two on each of its sides, and placed close to the angles (PL VI. fig. IG). 

 The number of segments in each cirrbus is eleven or twelve besides the terminal claw 

 (PI. III. fig. 10), and the width of the two basal segments somewhat exceeds their 

 length. In the third the length and breadth are nearly equal, but in the fourth and 

 fifth, wliicb are the two longest segments of the cirrbus, the proportion between them 

 is as about 4 to 3. The sixth segment is slightly shorter than the fourth, and from this 

 point to tbc end of the cirrbus the length of the segments gradually decreases, while at 

 the same time tbey cxbibit a slight degree of lateral compression. The dorsal spine, 

 tbe first indication of which is seen in the fifth and sixth segments, becomes very 

 marked indeed towards the end of the cirrbus, and develops in tbe penultimate segment 

 into a stout opposing process. 



(§ 40) The development of the cirrhi oi Act. poly morpha seems to take place somewhat 

 differently from their development in Aiit. rosacea as described by Dr. Carpenter '. In 

 tbe latter species tbe individual segments usually present all the characters of maturity 

 from a very early date, viz. the relative proportions in the length and breadth of the 

 segments, the bevelling off of tbe opposed faces on tbe aboral side, and the develop- 

 ment of the opposing process on tbe penultimate segment. But in some rare cases, even 

 after tbe cirrbus bas attained a considerable size and bas the normal number of seg- 

 ments, the latter are of a very rudimentary character ; their basal segments are the 

 longest, and tbe following ones rapidly decrease in diameter, so that tbc whole cirrbus 

 tapers considerably fro)n its base to its point. This condition gradually becomes less 

 and less marked as tbc segments increase iu size, and their opposed faces become 

 bevelled off towards the aboral side, so that tbe cirrbus ultimately acquires all tbe 

 characters of maturity. 



This mode of development, which is the exception in Ant. rosacea, seems to be the 

 rule in Act. polymorpha. All the very young cirrhi, both of the type and of Varieties 

 1 and 2, which I have met with, taper rapidly from tbe base to the apex ; and while 

 tbe four or five basal segments exhibit from a very early period tbe same proportion 

 between length and diameter as is seen in a completely developed cirrbus, the following 

 ones are still in a very rudimentary condition. The sixth segment, instead of being as 

 long as the fifth, is much shorter ; the seventh is still shorter and more slender, while 

 the terminal segments are little more than a succession of small disks ending in a small 

 and very rudimentary claw (PI. III. fig. 8 b). They are thus not only of the smallest 

 dimensions, but have a much more immature appearance than the basal segments; and 

 it would therefore seem that the augmentation in the number of segments is effected 

 by tbe interpolation of new segments, not at the base, as is usually the ease in Ant 

 rosacea, but between the middle segments and the terminal claw. 



' riiil. Trans, he. cit. p. 711. 



