38 
Becords of the Geological Surveg of India, 
[voL, xiir. 
slightly inclined, but which from their structure seem to correspond with the 
lower Eocenes of the Sasp)dl stream; their relations to the higher Eocenes are 
not, however, well displayed, and to the south-east they are concealed by a 
modern boulder deposit, which also covers the base of the Tertiary series near 
Leh, -where it has been much denuded away along the valley of the Indus. 
Above Leh, however, near the village of Arpa, we find the hard gneiss conglo¬ 
merate occurring low down in the Tertiary series, underlaid along the bed of the 
Indus by soft gravels, conglomerates, sandstones and clays with a south-westerly 
dip of about 15°. The conglomerate contains pebbles of a trap, which is thus 
shown to be of i«/ra-Eocene ago, but whose origin is not certain: the lowest 
conglomerate also contains pebbles of blue limestone and buff dolomite probably 
derived from rocks of the Carho-Triassic series, and indicating former outcrops 
of these rocks now probably concealed by the Tertiaries. Occasional blocks of 
gneiss, several feet in diameter, occur in the sandstones. From the softness of 
these rocks they have in great part been denuded away by the Indus, and only 
patches remain here and there. 
I can but think that ice-action has played some part in the formation of 
these lower Eocene strata, as it seems to me very difiicult to imagine that water 
power alone could have placed these blocks in their present position without 
scouring out the soft sand in which they are embedded. I have, however, no 
positive proof to bring forward in support of this view. 
I will now describe two sections taken aci-oss the nummulitics higher up the 
Indus than the former, the first being an ascending and the second a descending 
section. The first section is taken from the Indus valley below Leh to Kio in 
Zanskar. The lowest nummulitics exposed on the Indus near the village of 
Phay, consist of brown and green sandstones, mingled with coarse conglomerates 
and grits; the sandstones are often ripple-marked, and the pebbles in the con¬ 
glomerate consist mainly of gneiss, while the grits are composed of coarse 
gneissic sand precisely similar to that which at the present time is found in the 
valley of the Indus. In the higher part of the series, near Uriicha, purple and 
green shales and slates succeed and partly replace the grits, forming an anti¬ 
clinal resting upon green and brown splintery shales. The higher slates contain 
numerous bands of earthy limestone abounding in nummulites: the latter are 
particularly common near the village of Shingo,^ where I also obtained a species 
of Oonus. The rocks are here much folded, but the foldings are regular and open, 
and never show the minute contortions and crumplings so characteristic of the 
older rocks. Some 2 miles above the village of Kio the nummulitic rocks are 
overlaid by several hundred feet of a coarse conglomerate, which is here nearly 
vertical. The relation of this conglomerate to the shales is not very clear, 
but it aj)pears to lie in a synclinal axis, being again underlaid by colored 
shales nearer Kio; the great mass of purple shales are, however, unrepresented 
below the conglomerates at Kio ; close to the latter place the Tertiaries are under¬ 
laid unconformably by Carboniferous rocks. The higher Tertiary conglomerate, 
here and in other parts of the same line, contains numerous pebbles of the 
* The village of Shingo is placed in the Atlas Sheet II or 2 miles too near to Kio. 
