r8o Records of the Indian Museum. [VoL. XVIII, 
in their broader and thicker caudal peduncles, and this was pre- 
cisely the difference noted in our specimens, we assigned them 
provisionally to N. stoliczkae var. leptosoma, in which it seemed 
necessary to include the var. productus. 
On dissection we found that the alimentary canal of one of 
these specimens agreed with Herzenstein’s figure (op. cit., pl. 
viii, fig. 12) of that of N. stoliczkae. The alimentary canal of 
an individual of the same lot but belonging to the stenurus type 
differed considerably, as may be seen from fig 8. 
There seemed, therefore, at this stage in our investigation to 
be good grounds for considering the two forms, though occurring 
together, as specifically distinct. On examining the other speci- 
mens in the collection we found two (from a stream running into 
the Ram-T'so lake in Tibet) that clearly belonged to the stenurus 
type. These had been confused with N. lhasae, which had also 
been assigned to N. stoliczkae. We also found two specimens from 
Leh belonging to this (the true stenurus) type. 
We dissected one of the two Tibetan and one of the Ladakh 
TexT-FI1G. 9.—Pharyngeal teeth of Nemachilus stoliczkae from Seistan (xX 25). 
specimens of stenurvus—to find that in both the alimentary canal 
agreed with that of the individual of the leptosoma type from 
Seistan. Subsequent investigations proved that the structure of 
both types was variable in this respect. The one constant differ- 
ence that we could find between sfoliczkae (s.1.) and stenurus lay 
in the proportions of the caudal peduncle, and even these varied, 
as may be seen from our table of measurements, within wide 
limits. It does not, therefore, seem justifiable any longer to 
maintain stenurus as specifically distinct. The difference is neither 
sexual nor racial, but appears rather to be a true instance of 
dimorphism affecting both sexes. 
If this be so, the apparently discontinuous range of N. stenurus, 
which is recorded only from the mountains near the source of the 
Yangtse, from Scardo north of Kashmir and from Seistan, becomes 
explicable, for N. stoliczkae has the widest range of any member 
of its family in Central Asia. 
Another point to be considered is the status of the differ- 
ent varieties of N. stoliczkae recognized by Herzenstein (loc. cit.). 
We find it difficult in the large collection before us to assign some 
