22 THE HIGHLANDS OF CENTEAL INDIA 



area covered by Teak, to the total exclusion of the Sal. 

 The whole of this region is composed. of a trap overflow; 

 and all around it, as soon as the gigantic and lateritic 

 formations recommence, the Sal again entirely abolishes 

 the Teak. Again, within the area of the trap and Teak, 

 in the valley of the Denwa river, 150 miles west of the 

 furthest limit of the general Sal region, is found a solitary 

 isolated patch of the latter, occupying but a few square 

 miles. Here the Sal grows on a sandstone formation. It 

 is surrounded on three sides by trap rocks, and there it 

 entirely ceases, and is supplanted by the Teak as the 

 principal timber tree. But how to account for this small 

 and unimportant outlier of the great Sal belt ? To main- 

 tain our theory, some link to connect them together should 

 be found. I think that a hypothesis, much less extrava- 

 gant than many which are introduced into such arguments, 

 will do so. Towards the fourth side of the Sal patch in 

 the Denwa valley lies the great open plain of the Narbada, 

 into which the sandstone formation extends, and passes 

 on along with primitive rocks, and with little interruption 

 from the trap, right up to the main body of the Sal forest 

 at the head of the Narbada valley. The Sal, it is true, 

 ceases in the open Narbada valley, but so does all forest, 

 the country haAring been completely cleared and cultivated 

 for many generations. It is not then a very violent 

 assumption to suppose that the Sal forest at one time 

 extended down the Narbada valley as far as the Denwa, 

 and that, when the country was cleared, this little patch 

 alone was left securely nestled under the cliffs of the 

 Mahadeo range, in the secluded valley of the Denwa, into 

 which there was no road until within the last few years. 



These are strange facts. But it would be still more 

 strange if a corresponding distribution of animal life could 

 also be demonstrated. Something of the kind is really 

 almost possible. Equally with the Sal tree, several 

 prominent members of the Central Indian fauna belong 

 peculiarly to the north-eastern parts of India. These 

 are the wild buffalo {Buhalus Ami), the twelve-tined 

 " swamp " deer {Rucervus Duvaucellii), and the red jungle- 

 fowl {Gallus ferrugineus). All these are plentiful within 



