CHAP. 3.] CONCLUSION. 285 
eventually be of great benefit to every one. Hor years and years the 
supply of rain has been very small: within the last nine years, the 
Cumbum tank has never been full; and thus it might be said that 
there is little use in building tanks if we cannot get the water in a 
country so welladapted to hold it ready for distribution. A probable 
cause of this diminution of supply is, that denudation of the forests, or 
jungle-covered hill country, has been going on from time immemorial 
in the most reckless way. This ought more particularly to be guarded 
against in the country under description, for, as may be gathered from 
our aecount of the strata constituting those hill ranges, the soil is not 
a favourable one for forest vegetation. 
CHAPTER 3.— CONCLUSION. 
It now remains to be seen whether it is possible to correlate these 
` Madras rocks wi an her series that ar 
Attempt at correlation em viL y E s 
with other Indian forma- known in India; and it is unfortunate that we 
VE have only their lithology and order of superposi- 
tion to fall back upon as yet in this investigation. Nearly all writers on 
these rocks have tried to clear up this point, and have more or less 
been led to consider them as related to rocks of nearly lithologically the 
same kind occurring 1n Central and Northern India. 
The fact of the diamond occurring in certain strata at three different 
ER ther occurrence: of points of the country has been looked upon as a 
diamond-bearing strata.  oveat crucial test; only that too little attention 
was devoted to the finding out what particular series of rocks contained 
the diamond beds. This has, however, been cleared up lately; and when 
it is known that the diamond occurs regularly and always in one set of 
rocks in each area of a formation, whatever 16 may be, we are to a certain 
extent justified in looking on the occurretice of the diamond beds in 
distant and unconnected series of rocks as a point of resemblance or 
(865285 5 
