186 Records of the Indian Museum. [VoL. XVIII, 
of fibrous connective tissue is thin and incompletely differentiated 
and is not continuous across the dorsal muscles at the base. The 
central core has a more fibrous structure with smaller, ill-developed 
vacuoles. The blood-vessels are few and poorly developed. 
We have cut sections also, as already stated, of the adipose 
fin of a species of Glyptosternum (Sisoridae) from the base of the 
Nilgiri Hills for comparison. It would be out of place in the 
present context to discuss the structure of this fin in detail. We 
may say here, however, that an inner core of highly vacuolated 
tissue, closely resembling that found in the fin of Adiposia, 
occupies the centre of the structure and that its wall consists of 
two regions, the structure of both of which differs considerably 
from that of the homologous regions in Adiposia. Our figures (pl. 
xvi, figs. 5 and 6) and the explanation of them will illustrate the 
differences sufficiently for our present purpose. ‘These differences 
are so considerable that there can be little doubt that the adipose 
fin of Adtposia, though (like that of the Siluridae) derived 
from the posterior part of the primitive dorsal fold, has origi- 
nated independently, probably in correlation with the assumption 
of the habit of burrowing in the mud of bodies of water liable 
to desiccation, and there aestivating or hibernating until the return 
of the flood season. 
It seems to be clear, therefore, that the soft fin! of Adiposia 
is a highly specialized structure, but that it is fundamentally 
homologous with the posterior precaudal part at the primitive 
dorsal fold. 
The function of this fin in Adiposia is possibly a double one. 
It may act as a reserve food-supply for a voracious fish that must 
occasionally be deprived of food for considerable periods. It 
probably is also an accessory breathing organ, to judge from its 
copious blood-supply, of use when the fish is buried in damp 
mud. 
Relationships of Adiposia.—From what has been said above 
it is, we believe, clear that A diposia is closely related to Nemachilus. 
Its resemblance to the Siluroidea is probably more apparent 
than real, being due mainly to the persistence, doubtless secon- 
dary, of a post-larval character and its slight modification. We 
have no reason to think that Adifosia is an extremely primitive 
form, as would be the case if the persistence of this one juvenile 
character were accepted as evidence of direct affinities with the 
ancestral forms of both the Cyprinoidea and the Siluroidea, for, 
indeed, the dorsal fin-fold is an ancestral feature common to al 
fishes, and even to other groups of primitive vertebrates. In all 
fish with a dorsal fin of any kind part of it persists and the adipose 
fin of Salmonidae is not supported by other evidence as proof 
of close affinity with the Siluroidea. 
| A recent investigation of well-preserved specimens of Acanthophthalmus 
pangia proves the existence of a similar structure in that species. 
