DR. P. H. CAEPENTEE OX THE CIRRI OF CERTAIN COMATUL.E. 477 



joints, though the length is only 47 nam. (PI. LV1I. fig. 29), and in the most extreme form 

 from this locality there are 48 joints in a length of hut 35 mm. (PI. LVII. fig. 33). In the 

 Scotch variety therefore, while the number of joints is but little altered, they are relatively 

 very much shorter than in the Mediterranean examples. 



Let us now study the mode of development of these joints and the transitional forms 

 between one kind of cirrus and the other. Por the typical form of cirrus in the species 

 we must of course take the long-jointed ones which were described by Miiller in the 

 Mediterranean variety (PI. LVII. fig. 7). 



Type A. Long-jointed. — The three basal joints are wider than long, but the length 

 gradually increases from the first to the seventh and then remains tolerably constant for 

 some distance, decreasing slowly towards the end of the cirrus. But the terminal joints 

 ai'e not laterally compressed and wider than long, or even square, as is the case in the 

 cirri of most Comatnlae ; for they remain much longer than wide. In fact the width 

 diminishes slightly in the last few joints so that the end of the cirrus has a somewhat 

 tapering appearance, and there is a long terminal claw, without, however, any trace 

 of an opposing spine on the penultimate joint. Most of the cirri are of this type in all 

 the Mediterranean specimens of Antedon phalangium that I have seen ; and it is also 

 the prevailing one in those from off Cape Sagres and off Mondego. But in those 

 dredged by the ' Dacia' on the Seine bank it is very largely intermixed with another type 

 of cirrus (PL LVII. fig. 23). the terminal joints of which are rather short and broad, 

 somewhat as in those from the Boss-shire variety (PI. LVII. fig. 33). 



The ' Dacia ' dredgings, and in a less degree also those of the ' Porcupine,' yielded several 

 individuals in various stages of maturity; and I have therefore been able to study the 

 development of the cirri both in quite young individuals and also in those older ones in 

 which new cirri are continually being produced round the margin of the centro-dorsal. 

 So far as I have been able to make out, these typical long-jointed cirri are developed on 

 the rudimental plan. An early stage of their formation is shown in PL LVII. fig. 1. The 

 fifth and sixth joints are already distinctly longer than wide; but their successors 

 become rapidly shorter, and the rounded end of the cirrus is made up of a series of 

 annular joints without any trace of a terminal claw. If we make allowance for the meat 

 difference between the cirri of the two species, we may place this form between stages 

 C and D of the developmental series described by Dr. Carpenter in Antedon rosacea. 

 Its further development is shown in PL LVII. figs. 3 and 6. In tbose two cirri which 

 measure 14 and 22 mm. respectively, and especially in the latter, which has some fifty 

 joints, the lower joints present all the characters of maturity ; but the later ones are 

 mere discs, and it is only in the larger one (fig. G) that there is any indication of a 

 terminal claw and that but a very slight one. The claw becomes more marked in the 

 older cirri shown in figs. 4 and 5, and the later joints lengthen out till their shape more 

 nearly resembles that seen in the mature cirrus (PL LVII. fig. 7). As in the case of tin 1 

 rudimental cirri of Actinometra parvicirra l the addition of new joints seems to take 

 place in the outer half of the cirrus, those at its base presenting most of the characters 

 of maturity at quite an early date. 



1 Trans. Linu. Soe. 2nd ser. Zool. vol. ii. p. 60. 



