Uncifer {pu. in the same figures), a name chosen for the reason that in most species (but 

 not in S. orienialis, PL II. fig. 2g) it is deeply incised at the end on the inner side, so that 

 a strong hook is fiarmed. Pars media (PI. I, fig. la, w., fig. 2a, m., fig. 6^, w.) is the largest 

 and by far the most important portion of the petasma; it is connected with the peduncle of 

 the pleopod by a short piece projecting from the outer side near the base (see i. inst. PI. I, 

 fig. 7^); at the origin of that piece is found a process, processus basalis (PI. I, pb. in 

 figs. I a, 2 a, 3i:5, 6^, 7«) directed (on the figure) backwards and inwards, and differing in shape 

 according to species. Inside that process is seen an incision, and inside this the inner pro.ximal 

 portion of pars media, which has its pro.ximal margin incurved and in the main transverse. 

 More or less beyond the middle pars media is ramified, with a process and some branches or 

 lobes. I have attempted to find out the homologies of these ramifications in the diff"erent species 

 and have named them accordingly, because it is necessary to mention each process in the 

 description of the species. The first of these ramifications is named processus ventralis {^pv. 

 in the figures), as it originates on the posterior or lower side of pars media; it differs exceedingly 

 in shape and affords excellent characters, as may be observed by a comparison of pv. in the 

 following figures on PI. I: figs. \b and \c, 2b and 2c, ^c and ^d, \f and \g, "jd and "jc, 

 and it has no hooks excepting in 5. Gardiner i (figs. 2 b and 2c), while in S. seminudus (figs. 

 7(5 and ^ c) it is armed with thick, short spines, and in S. orientalis (PL II, fig. 2^, pv.) it is 

 quite small. The portion beyond the origin of processus ventralis is named capitulum, and 

 it has generally four lobes, in a few cases five or six. The most proximal lobe projects mainly 

 on the outer side and is named lobus armatus (PL I, la. in figs. \b and it, 2b and 2t, 

 3<5 and 3(-, "j b and jc, and more anomalous in figs. 4/ and ^g, 6/ and 6^); it has always 

 a number of hooks; the arrangement of these hooks and sometimes their size together with 

 shape and size of the lobe afford fine characters. The hooks on this lobe and on most or all 

 the other lobes show considerable differences in shape and extreme differences as to size, and 

 the membrane around them is generally invaginated, frequently to such a degree that the hook 

 is completely retracted. (In 1882 S. J. Smith published a good description with figures of such 

 hooks in Scrgestes robustus). The three other lobes are named respectively : lobus connec- 

 tens (/r.) which projects mainly outwards and has a single hook or generally a number of 

 hooks; lobus terminalis [It) which varies as to hooks, and lobus inermis (//'.) which 

 originates at the inner margin of the capitulum, possesses hooks in 5'. Gardmeri (figs. 2a and 

 2 b) but not in the other forms, and is wanting in 6*. Challenger i and S. fiilgens (PL I, figs. 4/ 

 and 4_f, figs. 6/ and dg). But in the two last-named species we find on the anterior side of 

 the capitulum near the origin of processus ventralis two lobes, lobi accessorii [lac. on the 

 figures quoted) which have no equivalent in the other species, and both are armed with hooks. 

 In S. orientalis and S. Edwardsii a. lobus accessorius is observed on the posterior side of the 

 capitulum towards its inner margin (PL II, lac. in figs. 2/1, 2k and 3a). 



In 1896 I divided the species of Sergestes into two groups according to the length and 

 thickness of the third pair of maxillipeds as compared with third pair of legs, and the equip- 

 ment with bristles or spines on their distal joints. With a small amelioration in the wording of 

 the second of the two diagnoses this division is still very useful, and is reprinted later on. 



