314 GAMMARID.1<;. 



last pair of caudal appendages seems to be a special 

 structure, having for its object the antenna-like use of a 

 delicate apparatus at the extremity of the body, in tlie 

 same way as the conical appendages at the extremity of 

 the body of the Orthoptera are developed in the mole- 

 cricket into a pair of elongated setose filaments. If the 

 Niphargus were simply a modified Gammarus — which has 

 two equal-sized branches to each of the same pair of 

 caudal appendages — one cannot understand why both the 

 branches should not have been equally increased, instead 

 of one being almost obsolete and the other immensely 

 developed as it is in Niphargus, or why it should obtain 

 an additional joint. It would be easy to say, that as the 

 mole-cricket has only need of two anal feelers, the Ni- 

 phargus requires only two, and therefore that one of the 

 branches in each pair of the Gammarus has become de- 

 veloped by use, and the other obsolete by disuse ; but as 

 the Niphargus has four frontal antennee, whilst the mole- 

 cricket has only two, the former should use four anal 

 feelers instead of two. We must here also allude to the 

 excellent manner in which Hosius combats the opinion 

 of M. Gervais, that this species is only a variety of his 

 Gammarus pvlex which has remained in a state of imper- 

 fection from dwelling in deep dark wells. 



Although we can find no fresh-water ally to this 

 genus in the rivers and streams of Europe, yet Bruzelius 

 has taken in the deep sea, near Bohusia, a form which 

 he has described under the name Eriopis elongata, 

 approximating so nearly to it that it appears to be 

 scarcely generically distinct. 



We are inclined to think that Gammarus pungcns of 

 Milne Edwards, from the warm springs of Cassini in 

 Italy, also belongs to this genus. 



