[voL. xin. 
17u Records of the Geological Survey of India, 
spreail, till their united waters discovei’ed the weakest part of the barriers enclos- 
ing them. This was at Dhansila, which at once became the point of dischai’ge 
for their joint waters, the constant cirtting action of which has resulted in the 
features the ground now displays. Whilst then the glacier of Bhim Tal was 
descending west of 13hansila, the glaciers descending from the heights between 
Mahi-agaon (behind Naukatchia Tal) and Padani, wei’e pushing straight across 
the ground immediately below them, with the result that the Padani glacier 
cut through the trap ridge, and initiated the channel by which the waters of 
both lakes esca]ie, whilst the Mahragaon’ glacier pursued a parallel couivse 
west of Sitalahat, its ‘moraines’ serving to impound the drainage between 
Padani and ifahragaon, thereby creating Naukatchia Tal. 
The point of supreme intei’est of course is, why a cii’cuitmrs exit across a 
hard trap ridge should have been selected by the waters of both lakes in jDre- 
fei’ence to a more direct and undeniably natural cotii'se across the apparently 
Aveaker obstacle of a landslip or ‘ boulder bed.’ As regards a boulder bed initi¬ 
ating a lake, I may observe that eA'ory Himalayan gorge is full of boAilders, 
but in no instance do they giAm rise to lakes. A flood may throAv a bar of them 
across the channel, but the next flood makes a clean SAveep of the obstruction. 
Of course the ‘ boulder bed’ which could be presumed to haAm given rise to the 
lake could not have itself originated therein, and besides the Kumaun lakes, from 
their limited dimensions and sheltered situation, do not produce hoiilders, being 
veiy different from those grand Italian lakes 
“ Tu Lari maxhne, tuque 
Fluotihua et fremiiu assurgens, Benace, marino,” 
and in the deepest of them, Naukatchia Tal, the sides may be seen dipping down 
at a steep angle and composed of angular fragments cAudently never disturbed 
by wave motion. The result of my own examination of the point of exit of the 
joint Avaters bcloAv Dhansila Avas very instructive. At Dhansila there is no doubt 
of the ridge Avhereon it stands being composed of rock in situ, but across the 
gorge through AA'hich the stream passes, that is, exactly south-east of Dhansila, I 
could detect no rock in situ. The entire ground is covered Avith loose subangular 
masses of hard trap, up to even 100 feet girth, but no rock in situ could I detect 
after a careful search. There are some rocks of course Avhich from their com¬ 
position decompose so freely at the sui’face, that it is not easy to find a clean 
natural section or exposure of them ; but such is not the character of the harsh 
intractable trap of Mahva Tal and the neighbourhood, such as formed the bulk 
of all the fragments here strCAved about. To me the conclusion was overAvhelm- 
ing, that 1 was on the ‘ moraine ’ of an old glacier which had Avound past Dhan¬ 
sila, and along Avhose channel the escape waters of the two lakes had made 
their passage. 
The ice Avork of the old glacier, sawing its way across the obstructive ridge 
at Dhansila, renders it easy to understand hoAV the AA'aters of both lakes, on 
their natural point of discharge being blocked, came to select this point as the 
» That above Naukatchia Wl, not the villags of tho same name, north of Bhiui Tal. 
