478 DR. P. H. CARPENTER ON THE CIRRI OF CERTAIN COMAIJL^. 



It sometimes happens that the terminal claw makes its appearance when the cirri 

 are much shorter than that represented in PI. LVIL fig. 5. An instance of this is 

 shown in fig. 2, and the later stages are seen in figs. 8 and 9. The claw is at first 

 almost central in position (figs. 6, 9), hut it gradually moves over towards the dorsal side, 

 though no opposing spine is developed, as seen in PL LVIL figs. 5 and 8. 



I have not found any cirri of this long, smooth, and very long-jointed Mediterranean 

 type among the Scotch individuals of Antedon phalangittm ; the nearest approach to it 

 being the cirrus, of which the end is shown in PL LVIL fig. 16. But this rather belongs 

 to the intermediate type (B) next to be described, more southern examples of which are 

 shown in figs. 15 and 17. 



Type B. Intermediate. — Besides these greatly elongated cirri with numerous joints, 

 the terminal ones of which are very long and slender, most of the specimens from the 

 Mediterranean and from the Seine bank have somewhat shorter and smaller cirri with 

 fewer joints, the terminal ones of which are relatively broader than in the typical form. 

 Of this kind there are many varieties ; while there is often a more or less distinct 

 opposing spine on the penultimate, and sometimes also on the previous joint as well. 

 Fig. 17 represents a moderately young one without an opposing spine, and fig. 19 

 another still younger in which it is tolerably distinct. In both of these cirri the joints 

 exhibit all the characters of maturity, although they are less numerous than in the 

 longer but very undeveloped rudimental form seen in fig. 6. Older cirri of this pattern 

 may either have no opposing spine (tig. 15), though the dorsal edge is very slightly 

 serrate, which is not the case in the type (fig. 7) ; or it may be present, sometimes with 

 another behind it (fig. 11). In this specimen the joints forming the outer third of the 

 cirrus are distinctly longer than wide; but in those shown in figs. 10 and 11 this is less 

 evident, and in the former the opposing spine is double. This form with broader terminal 

 joints sometimes occurs in the Scotch variety of the species (PL LVIL fig. 16); and it 

 thus presents an approach to the cirri with still shorter joints which are so characteristic 

 of this local variety (PL LVIL fig. 33). I have not found sufficient material for working 

 out the development of this shorter but still long-jointed form of cirrus ; but I think 

 that figs. 12, 13, and 18 represent some of its earlier stages. The second of these, which 

 is not half the length of the very immature form shown in fig. 3, came off the smallest 

 individual of this species which I have seen. 



Type C. Square-jointed. — Besides the more or less long-jointed cirri typical of the 

 Mediterranean variety, the specimens dredged by the ' Dacia ' on the Seine bank bore a 

 number of cirri much more like those characteristic of the Boss-shire form. The end of 

 one of these is shown in PL LVIL fig. 23. It has 57 joints which gradually decrease in 

 length from the middle of the cirrus onwards, and the short later ones are somewhat 

 compressed laterally, but about as wide as long. In this cirrus there are spines on both 

 the joints before the claw, while there is a third spine on the rather smaller cirrus shown 

 in fig. 24. On the other hand there may be only one spine (fig. 25), or none at all, as in 

 that shown in fig. 27, which also has a very greatly reduced terminal claw. But the 

 relative shortness of the joints in the outer third of the cirrus is very marked in all 

 these examples. 



