232 
Records of the Geolofjicnl Survey of India, [voL. Xiil. 
tlie remarkable tiling is, that if the Indus current brought this gravel over the 
top of what now form the Barukot ridge, it did not pass bodily round Gand- 
garh hills and flow through the plains east of Ganclgarh. 
The accumulation of ‘ erratics ’ rafted by ice, up a backwater of a tributary 
during a flood in the main river, could be easily understood; but not so a 
thick deposit of boulder gravel, unless on the supposition that Bie well-rounded 
character of the boulders is no bar to their being a veritable glacial deposit 
also. And this conclusion is, I think, the only one admissible. A strong current 
alone (if we dismiss the idea of ice flotation) could have accumulated the 
coarse cap of boulder gravel and clays here seen, between Dari and Barukot; 
but if the Indus flowed as far as Dari in full cuiTent, it must have flowed 
further towards Haripur ; but this it unquestionably never did, and the conclu¬ 
sion therefore is that this deposit is one formed by floating ice charged with 
clay and boulders, which accumulated in the ‘ cul de sac ’ of the Dohr mouth and 
deposited its burden on melting all over the area submerged by the Indus flood 
water. The reason why the Indus did not flow further into the Haripur plain 
is of course that the plain was then higher than now, and barred the further 
advance of the flood waters laden with ‘ en-atics ’ in that direction. As I have 
remarked, the boundary of this coarse boulder deposit is very sharp and defined, 
and this belt of gravel has served to check denudation over the area covered by it, 
while the irndefended alluvium of the Haripur plain has been greatly lowered 
by the scour of the Dohr and its feeders, the result being the scarped gravel-cap¬ 
ped ridge on which Barukot stands. Parts of this deposit have a certain resem¬ 
blance to the TaloJiir hotdder hed, in consisting of a compact clay in which boulders 
are disseminated, as may be seen in the cuttings on the Torbela road, though the 
bulk of the deposit is a coarse stratified boulder gravel. 
The following localities for ‘ erratics ’ may be specified. From Barukot to 
Toi’bela along the course of the Dohr, and on the left bank of the Indus for 
some miles above Torbela, some of these gneiss erratics are over 100 feet in girth. 
Hear Attock small erratics, as well as rounded boulders, are freely scattered about 
to a level of from 300 (or perhaps 400) feet, or thereabouts, above the present 
bed of the river. South of Haji Shah a gneiss ‘eri-atic,’ some 20 feet in girih, is 
lying in a field, and some smaller ones near it. One mile north-east of Camp- 
bellpur is a gneiss ‘ erratic ’ 30 feet in girth, and in the low ground near Campbell- 
pirr, an enormous number of ‘ erratics ’ fi-om 20 to 30 up to near 100 feet in 
girth. The largest of these are limestone, but some of the smaller ones are of 
gneiss. South-east of Campbellpur, near the village of Boliwala, a little orrtside, 
seemingly, the limits of the Indus bed, or Haro (Hurroh) channel as it now is, 
a few ‘ erratics ’ are seen 12 feet or more in girth ; and these blocks were pro¬ 
bably stranded there during floods. At IJ mile east-south-east from Boliwala is 
a granite ‘ erratic ’ in a field by itself, with ‘ cup marks ’ on the top. 
The whole country in this direction, towards Kala-ka-serai, is an old high 
level alluvium, co-extensive with the Indus alluvium of the lower Haro (Hurroh) 
valley, but no en’atics save those near Boliwala on its edge were noticed in it. 
Some 3 miles north-east of Choi, on the road fi-om Campbellpur, a huge and 
jagged erratic of quartzose gneiss occurs. It has spontaneously broken up into 
