GAMMARIDHA. 903 



culate edge, and the mandible itself has a denticulate summit. These 

 constitute the subfamily Stegocephalin^e. 



The remaining species, with the base of the inner antenna more 

 elongate and slender (the second and third joints of these antennae 

 being much oblong, instead of transverse), are naturally distributed 

 according to certain characters in the mandibles, the maxillipeds, and 

 legs. . 



The Leucothoin^i are peculiar in having the maxillipeds long and 

 slender, and the inner lamellar process very small or obsolescent; and 

 some or all of the species have no mandibular molar prominence. The 

 genus Stenothoe is of this subfamily, although peculiar in wanting the 

 mandibular palpus. Only the first and second pairs of legs are pre- 

 hensile. 



The Gammarinje, like the Leucothoinre, have hands only to the first 

 and second pairs of legs, and even these may be wanting. The man- 

 dible has a molar prominence, a doubly edged denticulate summit, a 

 three-jointed palpus. The maxillipeds are rather stout, and the inner 

 lamellar process is elongate. 



The Pontofobein^, unlike the Gammaridaa, have the legs of the 

 third and fourth pairs more or less prehensile ; and the J&MNM, have 

 the legs of the fifth, sixth, or seventh pairs prehensile. 



It is possible that a better knowledge of the characters of the man- 

 dibles in the Pontoporeinae and Isaeinae may require a different ar- 

 rangement of them. The genera of the former subfamily are known 

 to us only from Kroyer' s descriptions. 



Among the Gammhrince, there are two series of species, the Amphi- 

 ihoe series, having the superior antennae simple ; and the Gammarus 

 series, having these organs appendiculate. 



In the Amphitlioe series, the genus Amphit7ioe, as at present adopted, 

 includes all the species with the first and second pairs of legs subche- 

 late, the posterior legs of ordinary length, the mandibles with a molar 

 prominence. In Acanthonotus of Owen, the legs of the second pair 

 have no trace of a hand ; in OEdicerus of Kroyer, the posterior legs 

 are very long, and the claw is obsolete ■; in Michrocheles, Kroyer, the 

 mandible has no molar prominence ; in PJwtis, Kroyer, the fifth pair of 

 legs is described as reversed, and the last pair of stylets as having one 

 of the branches rudimentary. Other subdivisions have been proposed, 

 but they are not accepted by Edwards \ and, in fact, they were not 

 instituted on characters of importance. Pherusa of Leach, distin- 



