26 THE VOYAGE OF H.M.S. CHALLENGER. 



terminal part forms a very elongate five-jointed stem. Of the joints the carpal, as in 

 the maxillipeds, is by far the largest, being even longer than the whole of the preceding 

 part of the leg. It is greatly compressed and somewhat expanded toward the end, being 

 fi-inged moreover at the distal part of the exterior edge with a row of very long, 

 anteriorly curving, ciliated bristles. The inner edge of this joint, too, is likewise 

 provided wth several slender bristles, as also with a dense series of delicate curved 

 spiuules, crowded together at the distal part. The succeeding joint (propodus), which, as 

 a rule, along with the carpal joint forms a strong geniculate curve, is likewise rather 

 elongate, but considerably narrower, somewhat curved, and densely setigerous, more 

 especially at the inner edge. The terminal joint or dactylus is narrow, lanceolate, and 

 armed at the inner edge with a dense row of small spinules, as also with a few elongated 

 bristles. The exopod, as in the succeeding legs, is developed into a powerful natatory 

 l)ranch, on which can be distinguished a somewhat thickened and strongly muscular 

 basal part, together with a narrow and very flexible multiarticulate terminal part, 

 furnished on both edges with strong natatory setiB. At the base of this leg is attached 

 a fully developed gill of precisely the same structure as that characterising the five 

 succeeding pairs of legs. Projecting from the outer side of the basal part, may also be 

 observed a very small linguiform lobe, fringed with several exceedingly long and slender 

 diverging bristles. This lobe, which also occurs on the remaining legs (see fig. 10, ep), 

 would seem to represent a kind of rudimentary epipod. 



The succeeding legs (see PI. VIII. figs. 10, 16) are all comparatively uniform in 

 structure, and very similar in appearance to the first pair, described above, ditl'ering 

 only in the carpal joint being somewhat less expanded and without the long ciliated 

 bristles at the outer edge, and also in the propodal joint being straighter and having the 

 bristles arranged in more or less distinct fascicles. This arran<>ement of the bristles 

 induced the late Dr. v. Willemoes-Suhm to describe the terminal portion of the legs as 

 subdivided into short articulations, as in the- Mysidfe, a character which, however, has not 

 been proved in reality to exist. 



The legs, having all the character of true pereiopoda, as a rule diminish somewhat in 

 size posteriorly, likewise assuming, successively, a more slender form (see fig. 16). On 

 the last pair (fig. 14) the carpal and propodal joints become very narrow and of nearly 

 uniform length, both being furnished with numerous distinctly defined fascicles of 

 bristles. The terminal claw moreover is rather smaU and of a conical form. 



The gills (see PL VIII. figs. 16, 17) exhibit a rather complex structure, and occur 

 at the bases of all the legs. At the last pair, however (see fig. 14, br), they are very 

 small and rudimentary, merely consisting of a single branch placed at the outer side. 

 On the other hand, all the remaining gills (six pairs in number) ai-e of a perfectly 

 uniform structure, consisting of no less than four principal branches (see figs. 9, 11) 

 springing from a common base. Of these branches the largest, as in Lophogaster, is 



