Chap. 12.] w. blanford^ western india. 69 



now found. It appears scarcely probable that if, in addition to the 

 lapse of time^ so great a change in the physical characters of the 

 country should have taken place as is involved in the conversion of 

 a great series of lakes (a) into a simple river valley, no alteration should 

 have occurred in the fresh-water molluscan fauna, except the addition 

 of fresh species. Even the changes which have taken place are to some 

 extent susceptible of explanation. Thus Dr. Falconer, in a most valuable 

 paper upon the ossiferous deposits of various eastern rivers, remarks espe- 

 cially upon the absence, in the Nerbudda deposits, of Melania spinulosa, 

 Lam., and M, variabilis, Bens., and the rarity oi Planorbis Coromandelianus, 

 ■ Fabr. But M, variabilis, and also 31. lyrata, Bens., appear to be confined 

 at present in India to the Granges valley and the neighbourhood of 

 Bengal. M. lyrata, it is true, is mentioned by Mr. Theobald as inhabiting 

 the Nerbudda valley, but it is certainly rare, (5) and to the west it does 

 not occur. M. spimtlosa is very locally distributed throughout India. 

 The rarity of the common Planorbis is due to another cause. It is not, 

 as a rule, a river shell, it abounds in marshes, and the shells met with 

 in the rivers are chiefly dead shells washed in during floods. Such dead 

 shells, if light, are carried along upon the surface of the water and 

 stranded on the bank, where they decay. In the beds of Indian 

 rivers, by far the greater portion of the shells met with are species of 

 Paludina, Melania, Uiiio and Corbicula, precisely the species found most 

 abundantly in the deposits of the Nerbudda and Taptee valleys. They, 

 in consequence of their solidity, and the ease with which they fill with 

 sand, sink to the bottom, and Lymnea, which in consequence of its large 

 mouth, is also less suited for floating, often accompanies them. 



Both the nature of the deposit and its organic contents therefore 

 appear in favour of its fluviatile origin. It is also evident that if the 



(a). The upper Nerbudda alluvial plain exceeds 200 miles in length. The Berar plain 

 is about 110 miles long, that of Khandeish 160 miles, 



(6). I have nevex* seen it : I have met with both M, tuherctdata and M. spinulosa in 

 the Nerbudda. 



( 231 ) 



