240 Records of the Indian Museum. [VOrL. 2 valida 
The Cobitidae of Seistan, on the other hand, must be associat- 
ed with the Schizothoracinae in origin. Two of the three species 
belong to the peculiar genus Adzposia, otherwise only known from 
Turkestan, and are apparently endemic as species. ‘The third 
(Nemachilus stoliczkae) belongs to a group in its genus characteristic 
of the Central Asiatic highlands, and resembles its namesake of 
the genus Schizopygopsis in geographical range. 
The majority of the fish of Seistan are, therefore, without 
doubt of Central Asiatic origin and can only have reached Seistan 
from the northern watershed of the Hindu Kush, while a minority 
have probably arrived in the district from the lower parts of 
Baluchistan. 
From a geographical point of view, the molluscs are perhaps 
the most interesting group in our fauna except the fish. They 
differ considerably from the true Eurasian species that have pene- 
trated from Central Asia as far south as the valley of Kashmir, and 
almost as much from those characteristic of the Persian Plateau. 
This fact is illustrated equally well by the species and genera that 
are present and by those that are absent. The Seistan fauna in- 
cludes none of the widely-distributed Eurasian species found in 
Kashmir, such as Limnaea stagnalis and Bithynia tentaculata, 
nor does it include any representative of the essentially Eastern 
Palaearctic genus Melanopsis, common in Persia proper and 
Mesopotamia, or of Bullinus, one species of which is common in 
Mesopotamia. ‘The species of Limnaea that do occur bear a 
distinct resemblance to European forms, but at least one of 
them (L. bactriana) also resembles an Indian form, L. chlamys. 
The three species of this genus, one of which (L.hordeum) is 
very rare and is only known from empty and possibly sub- 
fossil shells have all been found also in Lower Mesopotamia, though 
not in Persia proper, but are not dominant in the former 
country. Two of them (L. bactriana and L. gedrosiana) also 
occur commonly in the hill-country of Baluchistan and Afghanis- 
tan, but not, so far as we know, at high altitudes. The three 
Planorbidae have a wide range both in the Oriental Region 
and in neighbouring districts. That of the two species of Gyraulus 
(G. convexiusculus and G. euphraticus, extends at any rate from 
Mesopotamia to Burma and all over the Indian Empire, while 
the third species of the family (Segmentina calathus) is found in 
Burma and Sumatra as well as in northern India. The occurrence 
of a species of Vzvipara in Seistan is an interesting feature. The 
genus is practically cosmopolitan, but for some unaccountable reason 
is absent from Syria, Palestine, Mesopotamia, the greater part 
of Persia and Baluchistan. Among living species the Seistan form 
(V. helmandica) is most closely related to one from Sind (V. sindica). 
It is, however, still more closely related to a fossil (tertiary) species 
from the Bugti Hills in south-eastern Baluchistan. Indeed, it 
can be separated specifically from the fossil form only with difh- 
culty. Both V. helmandica and the only Unionid known from 
Seistan afford clear evidence of the existence of an Indian element 
