AM I'll lIMili ( KUSI A( l.A in- I'lll'. CENTS CAMMAKUS 69 



margins, and so forth. Karaiiian stated tliat such characters as the longer dactvH and long 

 setae on the pleon somites show that the animals may have teen originated from a deep lake. 

 The Yale North India Expedition brought hack two samples of gammarids of Tso-Moriri 

 (L82), one being captured among the weeds of the estuarine water at the north end of the 

 lake and the other (only one individual) obtained by a vertical haul (30-0 m. ). These speci- 

 mens agree rather well with the typical Gammarus pidc.v in various characters, except the 

 third pleon somite which has a somewhat produced hind corner. The elongation of the 

 dactylus of each pereiopod as noted by Karaman is also seen in our specimens from Tso-Moriri, 

 not only in an animal obtained by a vertical haul, but also in many individuals collected among 

 the littoral weeds. In the other localities, e.g., Mitpal Tso (L 76), Khyagar Tso (L 81), etc., 

 relatively longer dactyli were usually observed only in the forms obtained by vertical hauls, 

 so that such a character seems to be peculiar to the forms of free-swimming life or of deep 

 water inhabitants, as Karaman suggested. In other words, it must be regarded as an 

 ecological character due to the peculiar nature of environments; it is not of specific 

 importance. 



With regard to Rivulo gamtnanis stuliczkac, Karaman has pointed out no clear differen- 

 tial diagnosis from the other allied forms, only writing as ", so jener des R. pulex oder 

 R. balcanicus mc\\i?, Gemeinsames" (p. 129). In some crustaceans like gammarids, which 

 show great variations in various body parts according to difference of environment, it is 

 a very difficult matter whether we are to recognize various forms with slightly different 

 characters as a separate new species or to treat all different forms as variations of a single 

 species. In regard to the various forms of gammarids from Kashmir, Ladak and Western 

 Tibet, as already discussed in the foregoing pages, it is my present opinion that it is best to 

 distinguish all the related forms as the peculiar local races or subspecies of Gammarus pulex, 

 thus avoiding the use of one or two new specific names for those forms. As Karaman 

 (1. c. p. 128) writes as "Oberflachlich einen R. pulex ahnlich," R. stoliczkae may also be a 

 local race (subspecies) of the pulex-se.v\ts of Gammarus (s. lat.), which has differentiated in 

 this region of Asia. Karaman, moreover, noted that R. stoliczkae may belong to the oriental 

 group of Rivulogammarus (1. c. p. 129). 



The genus Riznilogammarus has been used by Karaman for certain forms of the pulex- 

 series of Gammarus, including pulex itself. I have at present only very scant knowledge as to 

 the validity of this genus, because I have been unfortunately unable to see its original descrip- 

 tion. If this genus is adopted, all the forms here concerned may be placed under it. As far 

 as I am able to understand, it seems to me that it may be better to use the generic name 

 Rivulogammarus as a subgenus of the genus Gammarus. In the present paper, therefore, 

 with this in mind, I have referred all the forms to the genus Gammarus (s. lat.). 



3. DiSTRUujTioN .\N'n Its Limiting Factors 

 A. GKO(a<ArnicAi. Distriuutkjn 



So far as our present knowledge goes, Gammarus pulex (Linne) is the commonest 

 freshwater amphipod, being distributed over the Palaearctic region, from England to Japan 

 and some parts of North Africa, but not found in America. In the high altitudes in Asia, 

 Chevreux (1908) recorded this species from some parts of Turkestan, such as Lake Issyk-koul 

 (1615 m. above the sea), the Pass of Karacolum (2000 m.), Lake Tchatyr-knul (3200 m.). 



