274 N. ANNANDALE. 



been different in several respects from those now prevalent in the outer channel. 

 The variation in salinity of the water in particular was probably much less extreme 

 and the sandy bottom must have consisted of finer particles. Lime must also have 

 been abundant, as large quantities of kankar (nodular concretions of carbonate of 

 lime) are dug from shallow pits at the edge of the lake, close to the place where the 

 shells are found. 



The small size of the shells of Meretrix from Barkuda shows that the molluscs 

 cannot have been living in particularly favourable conditions, but as they are less far 

 removed from the forma typica, which is the form that lives in conditions most 

 nearly normal for a marine mollusc, they possibly represent an earlier stage in the 

 evolution of the lake. Their dwarfed condition may be due to an admixture of 

 alluvial mud with the sand in which they once lived, and this also may have 

 had some effect in reducing the thickness of their shell, for heavy-shelled bivalve 

 molluscs are not suited to live in dense sticky mud, such as that of the main area of 

 the Chilka Lake. 



The light thrown by palaeontology on the history of Barkuda is not a strong one. 

 It is, however, sufficiently bright to show that the island is of recent origin as such, 

 and in considering its biology we are not concerned with the age of the ancient rocks 

 of which it is composed. The existence of deposits of shells of a distinctly estuarine 

 type, 30 or 40 feet above the present level of the lake, shows that its level has sunk 

 considerably since the lake was an open bay, and Barkuda must have been 

 completely submerged at the time at which the Rambha beds were laid down. 



III. VEGETATION. 

 General Character of the Vegetation. 



The flora of Barkuda, as might be expected from its scanty soil, infertile rocks 

 and by no means excessive rainfall, is of a restricted kind. A rigid natural selection 

 has taken place and many species, even genera and families, common on the main- 

 land a few miles away, are absent or maintain a precarious existence as stray plants. 

 The vegetation, however, is not a normal desert vegetation, for it consists mainly of 

 large trees, shrubs and thick-stemmed creepers and is, at any rate in the rainy 

 season, luxuriant in appearance. Neither is it a true equatorial jungle, many of the 

 elements of which are entirely absent. There are no epiphytes (except "parasitic" 

 figs), no tall grasses, reeds or bamboos, no palms, screw-pines, aroids or orchids, few 

 true parasites, ferns, mosses or liverworts ; and it is only where man has interfered 

 that herbaceous plants (with very few exceptions) have been able to gain a footing. 

 Moreover, the greater number of the creepers lose their leaves in the dry season. 



Nevertheless, in spite of it . restricted nature, the vegetation of the island can on 

 analysis be divided into several distinct types occupying different areas. The greater 

 part of the surface has been colonized primarily by a ficetum or fig-jungle in which 

 the dominant tree is the wide- spreading Banyan, Ficus hengalensis. In this ficetum 

 there is a fairly dense undergrowth, in which Glycosmis f>entaphylla is the most 



