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



from these, in the coxal joint wanting the luminous glohuk', and iu the terminal joint being 

 simply linear, and furnished with similar l:)ristles to those on the preceding joints. 



The three succeeding pairs of legs successively diminish in length, hut exhiljit other- 

 wise precisely the same structure as the second pair. The last of these pairs, or the fifth 

 in the series (fig. 19), has the joints somewhat more slender, and generally so bent that 

 the terminal part, when the legs are extended, curves in an opposite direction. The gills 

 attached to these legs are much more complex in structure than is the case with those on 

 the preceding pairs. 



(3f the two last pairs of legs no trace can be detected exteriorly. Only on dissecting 

 the animal and separating the two posterior pairs of gills (see PI. XL figs. 9, 11) does a 

 minute non-articulate stem, apparently the rudiment of the leg, become perceptible, 

 affixed to the inner side of each gill. This stem (fig. 12) is provided with a few* simple 

 bristles, and would seem to represent the endopod rather than the exopod. 



The gills (see PI. XI. fig. 5) are true " podobranchiiB," l)eing attached to the outer side 

 of the coxal joints of the legs, and thus, apjiareutly, i-epresenting the modified epipods. 

 There are seven pairs, wholly uncovered, as stated above, by the carapace, projecting at 

 some distance beneath its inferior margin, and arranged in a dense series along each side 

 of the trunk, partly overlapping each other posteriorly. They continue increasing in size 

 posteriorly, and the last pair are very much larger than any of the others, and partly project 

 along the sides of the first caudal segment. As to their structure, the four anterior pairs 

 (see PI. XII. figs. 16, 18, 20, 21) are much simpler than the three posterior, consisting 

 merely of an interiorly and anteriorly curving stem, from which issues posteriorly a 

 regular series of slender, digitiform, or filiform appendages, diminishing gradually in length 

 towards the apex, which appears more or less curled up. These appendages, representing 

 the true gill elements, exhi1)it internally, iu spirit specimens, a fairly regular double series 

 of small globular corj)uscles (fig. 22), which, ap});nvutly, are lilood-cclls, arranged accord- 

 ing to the centrifugal and centripetal course they take through the appendages during 

 life. The fifth and sixth pairs of gills (PI. XL fig. 9 ; PI. XII. fig. 19) are divided into 

 three branches, the two outer of which exhibit precisely the same structure as each of the 

 anterior gills, while the inner branch is distinctly bipinnate or furnished with a double 

 row of gill appendages. This br;inch, too, being the largest, is, as in the Loiihogastrid^e, 

 bent in beneath the trunk, meeting the corresponding branch on the opposite side in the 

 median line. Finally, the last pair of gills (PI. XL fig. 11) is far more complex in 

 structure than any of the others, the outer Ijranch being very large, and more or less 

 richly arborescent; or it may send off numerous secondary luanches, each of which pre- 

 sents a similar structure to that of the anterior eills. 



In what manner the ova, immediately after lieiug discharged from the ovaries, are 

 carried by the females of this genus, I am unable to state, none of the .specimens 

 examined having been furnished witli external eorp-baofs. 



