OF THE PARALLEL ROADS OF LOCHABER. 
701 
with destructive speed. The Glen Roy lake having an established outflow into Glen 
Spey, there could be no rise in the water-level, but the slow weathering of the detrital 
barrier—or, what is more probable, the gradual melting of the ice-barrier—of Glen 
Glaster pass would be sure there to lead to, sooner or later, a catastrophe of this nature. 
With what rapidity a small breach in detrital barriers enlarges and leads to such a 
result, was well exemplified by the Holmfirth flood. In that case, a reservoir with 
a superficial area of 11 acres, and 70 to 80 feet deep in the centre, had been formed 
by throwing across a valley a barrier 87 feet high, 16 feet wide at top, 480 feet at 
bottom, and 340 feet long, constructed of clay with a facing of gravel and stone.' 1 ' A 
small breach was accidentally formed, and so rapidly increased, that the barrier 
suddenly gave way, and a portion, 140 feet wide at top and 25. feet at bottom, was 
carried away at one swoop, and the reservoir drained in a very short space of time. 
Similar rapid bursting of reservoirs, under like circumstances, have occurred in the 
neighbourhood of Sheffield, in works connected with the Crinan Canal, and elsewhere. 
In cases of slight overflow of river banks like results have followed, and the inevitable 
consequences of such accidents are well known to all engineers. 
Nor, although we might fairly predicate what would be the effect of the bursting of 
an ice-barrier, are we entirely without evidence of the fact. In the first class of 
glacier lakes before described, the water escapes, sometimes quickly, at other times 
slowly, through fissures and crevasses in the glacier, which itself does not materially 
suffer. In the second class of larger lakes, formed by a glacier barrier transverse to 
the mam valley, the barrier is carried away suddenly whenever the pressure becomes 
too great or an overflow sets in, and the water is discharged with resistless rapidity. 
One of the best-known cases of such casualties is that recorded by Charpentier,! in 
which, in 1818, a glacier descending from a lateral ravine crossed the valley of Bagnes, 
and dammed back the small stream of La Drance. This soon formed a lake, which 
grew to be about 2 miles long and more than 200 feet deep at its lower end. In the 
summer the dam of ice gave way, and in twenty minutes the whole volume of water, 
estimated at 600,000 cubic toises, was discharged. 'The ground covered by the lake 
suffered little or no change, except near its outlet; but the torrent spread destruction 
as far as the valley of the Rhone. 
In 1844. Dr. Falconer'! drew attention to a great flood which took place in the 
valley of the river Lundaye (a tributary of the Indus), and swept away innumerable 
villages ; while at Attock, on the Indus, that river rose 30 feet above its usual level. 
The flood was supposed to have been caused by the rupture of a glacier from a lateral 
ravine, crossing the valley and damming up its head waters ; but Mr. Drew, who has 
since visited the district, found that the barrier had been formed lower down, and on the 
* Quart. Journ. Geol. Soc., vol. viii., p. 225. 
t ‘Essai sur les Glaciers,’ p. 204; and Reclus, ‘ La Terre.’ 
+ “ On the recent Cataclysm of the Indus,” Journal of the Asiatic Society, vol. x., part ii., p. 615, 
