94 
[VOL, VIII. 
Records of the Geological Survey of India. 
The section in Sikhim, to the east of the Nepal territories, is petrographically very 
different from that in the Simla region. The schists and gneiss come close to the edge of the 
mountains, in some places quite up to it. But in many parts there is a narrow hand of 
partially altered strata, which have been fully identified as belonging to the Damuda formation, 
th e coal-measures of India, thought to he of upper palaeozoic age. But the curious point is 
that these beds are the lowest and oldest member of the rock-series, conformably underlying 
the schists, and these again passing regularly beneath the gneiss, which forms the greater 
part of the mountains in the Darjiling district. In Sikhim the usual outer ranges of the 
sitb-Himalayan hills, enclosing long valleys at the base of the mountains, are not represent¬ 
ed. The inner zone of sandstone at the very base of the mountains is, however, in-force. 
In the Nepal section wo find a very complete exhibition of the sub-IIimalayan hills, as 
known to the north-west. The Churia Ghati range, in size, structure and appearance is a 
fac simile of the original Sivaliks. Inside it the dun or mari of Etoundah is an excellent 
example of these characteristic sub-Himalayan valleys. And to the north of this, along the 
Base of the mountains, there is a flanking range of sandstone, harder, and apparently 
older, than that of the outer hills, just as occurs in the western region. There is still so much 
uncertainty about the grouping and distribution of these sub-Himalayan rocks, that I cannot 
speak confidently as to those in the section under notice. It is quite recognised that there are 
two strong and stratigraphically well separated groups—the Sivalik and the Nahan in the 
trans-Jumna region; but considerable doubt has been thrown upon the view I at first adopted, 
that the cis-Jumna Sivalik hills belong to the upper group of rocks. Lithologically, 
the resemblance is more with the Nahun than with the Sivalik group. Thus it would 
appear that structural position, even in the case of what is physically a single range, is no 
criterion of the geological horizon of the rocks ; and wo are unable on these grounds to 
assume that the Churia Ghati strata are true Sivaliks. Lithologically too, they have small 
resemblance to the typical Sivaliks of the trans-Jumna range. In mere composition they 
are much more like rocks of the cis-Jumna hills, consisting as they do in the lower 
half, of massive gray sandstone, and above of great beds of conglomeritic gravel. There are, 
however, some points of difference: in the west the change from the sandstones to the con¬ 
glomerates is gradual and alternating; here it is rapid and complete, from an almost unbroken 
mass of fine grey sand to an equally uniform mass of pale yellowish-brown conglomerate. 
This character can have no significance ; but I was much struck with the very fresh aspect 
of these Churia Ghati deposits as compared with those of the range south of the Dehra-Dun. 
The sand, in solidity as well as in appearance, is scarcely different from that forming the 
chars (temporary islands) in the great river beds. I should, perhaps, mention that it is 
several years since I have seen the Sivalik sections, ^nd have since then been occupied with 
much more ancient formations. On one point, however, I will speak firmly: I must at 
present refuse to believe that the Churia Ghati strata can be of the same horizon as the 
sandstone forming the hills north of Etoundah ; and so, these being presumably Nahans, 
the former may for the present be set down as Sivalik. 
At the outer base of the range, at Bichiakoh, there are some rusty earthy beds ; and 
all are greatly crushed, locally quite vertical. The dip soon settles down to 30°, to north- 
north-west, maintaining it steadily to the top of the pass. This is the type structure of 
these detached sub-Himalayan ranges, of whatever group composed: the flat, hall of a 
normal anticlinal flexure. The range is about four miles wide, which would give an ag¬ 
gregate thickness of about 10,000 feet of rock* The pass, as is universal in these ranges, 
* It is still necessary to note that this does not imply vertical sequent' . 
