Prof. T. G. Bonney—Rounding of Alpine Pebbles. 55 
perimental researches of Daubrée in his “ Géologie Experimentale ” ;* 
but from neither these nor other sources (so far as known to me) could 
I obtain what I wanted. I must, however, admit that I am less 
familiar with the talus heap of geological literature than I should 
be, and prefer making observations in the open air, to hunting for 
records of them in a library. So, as I had some opportunities of 
doing the former during my journey last summer, I record the results 
in the hope that they may be useful to others; first heartily thanking 
my companion, the Rev. EH. Hill, for constant co-operation and as- 
sistance. 
These observations may be arranged roughly in three groups, 
which correspond with the three stages in the physical history of an 
Alpine river. It begins as a series of torrents, born for the most 
part high up on the mountain side from snow-bed and glacier. Its 
next stage is that of a single torrent rushing over the bed of an 
Alpine valley, still bounded on either hand by mountain ranges ; the 
third, and last for our present purpose, begins as it issues from the 
‘gates of the hills’ and commences its journey through the lowland 
plains. In the first of these stages its fall is always rapid, as it 
leaps from ledge to ledge, or even from crag to crag, descending not 
unfrequently at average rates of 250 feet to 500 feet in a mile, or 
along slopes averaging from 3 to 6 degrees. Thus it is a rushing 
roaring cataract, able always to sweep along boulders full a couple 
of feet in diameter, and often much larger blocks. In the 
second stage the fall becomes more gentle, though occasionally, es- 
pecially at the commencement, there may be a partial return to the 
former conditions; tributary streams also are being constantly re- 
ceived, which have only passed through that stage; but gradually 
as the valley widens the torrential character is lost, and the river 
flows as a strong swirling stream, in which intervals of actually 
broken water are rare. Lastly, after emerging from the mountains, 
and thus being cut off from all further contributions from torrents, it 
sweeps along with a strong steady flow, “hasteless but restless,”’ 
perhaps one of the grandest representations of unobtrusive power 
that can be seen in Nature. 
Obviously in observations of the present kind, little more than 
general results can be given. A river is constantly receiving 
tributaries, which discharge into it materials, not only differing in 
hardness or tenacity, but also in the amount of detrition which they 
have undergone. The strength also of its current varies from time 
to time. In some districts, or at certain seasons, what in the 
morning was a dry stream-bed, in the afternoon may be a roaring 
torrent. Further it is extremely difficult to determine the velocity 
with which a mountain torrent flows. In the case of some of the 
larger and less rapid rivers we made rough estimates by watching 
floating bits of wood and the like; they varied from about 21 to 4} 
miles an hour. An approximation also may be made from observing 
the size of the pebbles in the bed. A table is quoted in most text 
1 Vol. i. sec. ii. ch. i. 
