AMi'Mii'dii ('i<i!S'ia<i:a n|- 'rill-, cknus cammauils 73 



yVll these speciiiK-ns, i-xcei>t tlinse frdiii the lasl lncahly, apiiear l<i ha\e a pnjlonged 

 and pointed hind corner to the third pleon segment and an elongate telson : the Gammarus 

 from the spring at Ciuishol, on the other hand, exhibit indications of the characters of the 

 specimens from the southern part of Indian Tibet, particularly in their shorter telsons. 

 These records, therefore, confirm Dr. Ueno's findings based on the major part of the 

 collections. 



The following additional ecological evidence, moreover, supports some of Dr. Ueno's 

 conclusions. 



The oxygen content of the immediate environment of CaiiiiiKinis in the high regions of 

 Indian Tibet lay between 5.6 and 7.7 mgrms. per litre. In a stream at Dras, in the inter- 

 mediate zone and so free from Gaminarus, it varied from 6.2 to 6.7 mgrms. per litre; and 

 in a swamp at Spithug, a very favourable habitat in the intermediate zone, again without 

 Gamiiuinis, the water in contact with weed was probably supersaturated by day, though no 

 figures relating to this locality are forthcoming. 



Gravimetric analyses of the waters of some of the lakes have been made by Mr. J. A. 

 Newlands of Idartford, Conn., and it is now possible to state that the calcium content of the 

 Gammarus localities in Indian Tibet varies from 9 parts per million (Ororotse Tso) to 303 

 parts per million (Panggong Tso). Since the water of the former locality had the lowest 

 alkali reserve encountered (0.0003 N.) it is safe to assume that the waters of the intermediate 

 zone (alk. res. 0.0007 to 0.0036 N.) contained sufficient calcium to support a population of 

 Gmtimarus. 



Potamogeton pcctinatiis Linn, and Ranunculus triclwpliyllum Chaix (kindly determined 

 by Dr. E. D. Merrill and his staff at the New York Botanic Garden) were the dominant 

 species in such of the high Gammarus localities as contained higher plants, but more often, as at 

 Ororotse Tso, Togom Tso, Khyagar Tso, Panggong Tso, etc., no higher vegetation occurred. 

 Both P. pectinatus and R. trichopliylhim as well as charophytes occurred in suitable localities in 

 the intermediate zone. It is therefore clear that there is no correlation between the vegetation 

 and the presence and absence of Gammarus. 



Dissection has shown that the pelagic Gammarus of Khyagar Tso is apparently pre- 

 daceous, feeding on the large, dark, slow-swimming Cladoceran Dapliniopsis, the deep sepia 

 lirown chitin of this form being easily recognizable in consideralile quantity in the faecal 

 matter at the posterior end of the alimentary canal of Gammarus from a vertical haul taken 

 in this lake. Dapliniopsis occurred nowhere in the presence of fish (e.g., Pangur Tso, Yaye 

 Tso) though some lakes where no fish were found (Ororotse Tso, Tso-Moriri) lacked also 

 Dapliniopsis. In \ iew of these finflings, taken in conjunction with the results of other 

 workers who have studied the relation of Gammarus to fish, quoted above by Dr. Ueno, 

 there can be little doubt that the abundance of Gammarus in the closed and mineralized waters 

 of these Central Asiatic lakes is due largely to the absence of fish, which leaves Gammarus 

 unmolested on the nne h.and and on the other introduces no competitor for the most con- 

 sincuous source of food of these Amijhipncjs. — G. E. II. 



