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



apart from all the rest. The Lysianassidse are set apart in another direction by the 

 peculiarities of the upper antennae, the second gnathopods, and especially that character 

 of the mandibles, on account of which Schiddte has named them the Trochalognatha. As 

 far, however, as the antennae are concerned, they are united to Schiodte's other group, 

 the Eleutherognatha, by the new family Valettidse. The Stegocephalidse, while agreeing 

 with the Lysianassidse in the upper antennae, are less remote from Gammarus in the 

 form of the second gnathopods. On the other hand, the character of the mandibles shows 

 a further departure from the common type than is found anywhere else among the 

 Gammarina. Hence a common ancestry may be supposed for these three families, 

 branching ofl" from Gammarus at a remote period. 



In the Stenothoidse the genus Stenothoe itself, being without the mandibular palp, 

 may be regarded as a later form than the companion genus Metopa, in which that palp 

 is retained. In the Leucothoidse the genus Leucothoe, by the characters of its mandibles 

 and maxillipeds, seems to lead up to the Stenothoidae. 



The Syrrhoidae and Synopidae on several accounts may plausibly be placed side by 

 side. In one particular, the very short terminal joint of the mandibular palp, these 

 families show an affinity to the Stenothoidae ; but apart from this one point their 

 affinities are with the Pontoporeiidae. Nearer than any of these to Gammarus stand the 

 Gldiceridae. 



The Iphimedidae, Pardaliscidae, and Amphilochidae remain, with peculiarities that 

 make every suggestion for their classification hazardous. At a venture the Pardaliscidae 

 may be grouped with the Syrrhoidae, Synopidae, and Ponto^joreiidae ; the Iphimedidae 

 with the Pleustidae and Epimeridae ; and lastly, the Amphilochidae left, where they are 

 usually placed, in a somewhat dubious proximity to the Stenothoidae. 



Between the Gammarina and Hyperina there is a wide gap, over which at one point 

 no bridge has yet been found, for, while in the Gammarina the maxillipeds always have 

 a palp, they never have one in the Hyperina. In the Gammarina the mandibular palp 

 has, with the rarest exceptions, a short first joint, whereas in the Hyperina this joint is 

 frequently of great length, but here there are all sorts of connecting links, the maudil:)ular 

 palp in Cyllopus being quite of the pattern common among the Gammarina. Milne- 

 Edwards, in 1840, when establishing the Tribu des Hyperines Gammaroides, went so far 

 as to say that the single genus, Vibilia, which he placed in it as a link between the 

 Gammarina and the ordinary Hyperina, might almost as well stand in one division as the 

 other. To this overstatement of the closeness of the tie between the two groups he was 

 no doubt led by wrongly supposing that Vibilia had rudimentary palps to the 

 maxillipeds. 



Within the Hyperina, although marvellous diversity of form has been arrived at, there 

 is comparatively little difficulty in tracing a family resemblance between the difi"erent 

 sections. Naturally the Platyscelidae or Hypdrines anormales, with their strange zigzag 



