EEPORT ON THE AMPHIPODA. xvii 



world without the competition of other species of Amphipods. In order to enable the 

 family to extend its range over the fresh-waters of the world, no further change was 

 needed than such as would enable some of the progeny to pass from salt-water to brackish, 

 and from brackish to fresh. But the sections of this genus having once obtained command 

 of the rivers, by the capacity of living vigorously in the river-water, would have an 

 immense advantage over all rivals attempting in the future to make a lodgment in the 

 streams, while their capacity for life therein was in its initial stages and only feebly 

 developed. 



In the arrangement around and near to Gammarus of such genera as Niphargus and 

 Gammaracanthus and Mwra, there is in point of fact a very general agreement, so that 

 we shall not be very rash in regarding the Gammaridse as a natural family. If from the 

 considerations already mentioned we may regard it also as the typical family of the 

 Amphipoda, the next point of interest will be to determine whether the other families 

 can only be grouped confusedly around it, or whether any lines of succession can be 

 suggested. It is evident that if the Gammarus-fovvn had at any time such possession of 

 the ocean- world that it was able to stock the majority of the fresh- water streams with 

 genera and species which can be traced back to it in a direct line, the more or less 

 amphibious Orchestidas ought to be traceable to the same source. Between Gammarus 

 and Talitrus there are, it is true, important distinctions, but they are in part bridged 

 over by the genera Hyale and Hycdella, and to a very considerable extent they show 

 adaptation on the part of the Orchestidse in general to iiltered conditions of life. 



The next family which seems easily derivable from Gammarus is that represented by 

 the genus Amjjhithoe. It will be noticed in the history of the subject, that, before the 

 minute subdivision of genera, the discoverer of a new species, if he did not assign it to 

 Gammarus, was very apt to call it an Amjyhithoe. Amphitlio'e indeed has before now 

 been chosen for description as a typical genus of the Amphipoda. The genus, in its 

 present acceptation, has an extremely extensive distribution, and by the simplicity of 

 structure which it exhibits, and its general approximation to Gammarus, it is well fitted 

 to be the link between that genus and the nest-forming Podoceridse in general. Near 

 to, yet not to be derived from, the Podoceridas, and by their somewhat more erratic 

 characters placed at a greater distance from Gammarus, come the Photidae. Again, at 

 various distances beyond the Podoceridse, we may imagine positions for the Corophidfe, 

 Cheluridse, Dulichidas, and Icilidse. The Dulichidae seem to lead on by a very natural 

 sequence to the Caprellidse, with which Bate and Westwood have gone to the extreme 

 length of actually classing them, in their group Aberrantia. 



Eeturning to the family of the Gammaridse, we find in close alliance with it two 

 other families, the Atylida^ and Eusiridse ; nor are these remote from the Pleustidse and 

 Epimeridse. The remainder are by no means easy to group in any plausible order of 

 relationship to the ancestral form. The eyes of the Ampeliscidse give them a position 



(ZOOL. CHALL. EXP. — PART LXVII. 1888.) XxX C 



