1920.] N. ANNANDALE & S. lL. Hora: The Fish of Setstan. 155 
integument in rapid-running water. Similarly with the Cobitidae, 
which have probably lost their scales in acquiring the burrowing 
habit. But the fact that the Cyprininae also of Seistan are, as it 
were, casting off their scaly garment and by a different process 
from either the Cobitidae or the Schizothoracinae, suggests that the 
phenomenon has some other, strictly local significance, and that 
there is something in the environment of these fish that renders 
scales an encumbrance rather than a protection. But what this 
something is, we do not know.! 
Another general peculiarity of the fish of Seistan, possibly 
correlated with the degeneracy of the scales, is the brittleness of 
their fin-rays. ‘This feature isso well-marked that difficulty was 
experienced in preserving specimens with the caudal and dorsal 
fins intact. Possibly both phenomena may be due, directly or 
indirectly, to the peculiar composition of the water in which these 
fish live ; but this is a mere suggestion. 
The species all seem to be mainly bottom-feeders, with at 
least partly ventral mouths and more or less flattened ventral 
surfaces. They do not, however, possess any highly specialized 
tactile organs, and their eyes, though rather small, are not de- 
generate. ‘The fins are small, but at any rate in the Schizothora- 
cinae and Cobitidae, much larger proportionately in the young 
than in the adult. 
This is all we can say about the structural peculiarities of the 
fish-fauna of Seistan as a whole, but in two of the three species of 
Cobitidae a remarkable peculiarity occurs, namely, the persistence 
of the posterior part of the primitive dorsal fold in the form of 
a soft or adipose fin. This peculiarity has not been commented on 
hitherto in any Cyprinoid fish. It is not, however, found only 
in species from Seistan, for it is figured, apparently without com- 
ment in the Russian description, by Kessler in his Nemachilus 
longicauda from Turkestan. Moreover, as we will demonstrate 
later, the soft fin in these fish differs little in fundamental structure 
from the fold present in a young post-larval stage in the allied genus 
Nemachilus. Its persistence and slight modification in the species 
to which we give the generic name Adzposza is probably correlated 
with the necessity of burrowing in the mud in periods of drought. 
We will discuss the homology and function of the structure in 
detail when describing the genus. 
All we can say, therefore, on the subject of structural modi- 
fication in the fish of Seistan is that they are in several instances 
specialized forms, but that apart from a certain degeneracy of 
the scales, their specialization is not the result of evolution 
in their present home, but of long anterior specialization in 
the mountains of Central Asia. Their migration to the swampy 
t A suggestion has been made to us that the disappearance of the scales may 
be correlated with increased necessity for respiration by means of the skin, but 
this could hardly be affected by deciduous scales, which are only lost when the fish 
suffers rough treatment. 
