68 Proceedings of Royal Society of Edinburgh. [sess. 
caused largely by what may be described as a nearly horizontal hail- 
storm of small stones. Some idea of the density of this storm was 
obtained by counting the number of marks on a measured area. 
This was found to be from 250 to 300 per square foot, and it must 
be remembered that most of these were probably made after the 
bark had been removed by the first part of the storm. At a distance 
of 3 miles from the crater the bark had been stripped from only 
very young trees, but the marks of blows could easily be seen on the 
bark of all the trees. 
In following the earth-torrent from Mine up the bed' of the 
Biwasawa, I was immediately struck by the extraordinary effects 
produced by erosion in the short space of ten months which had 
elapsed since the eruption. The Biwasawa is only a small stream, 
and yet in these ten months it had carved out a valley which at 
one place was, by actual measurement, 80 feet deep and 80 feet 
wide at the top, and at other places, where measurement was 
impracticable, was by estimation little short of 150 feet deep. It 
must be remembered, too, that this one cutting does not represent 
nearly the whole of the work done by the stream, which had not 
flowed along the same channel all the time. Had the stream flowed 
quietly along its channel for the whole of the ten months, it is 
probable that it could not have made nearly so deep a cutting, but it 
was evident that much of the work had been done by a succession of 
floods. In one place, for instance, a lake had been formed which, 
after reaching a large size, had burst through the embankment which 
held it back, and the water descended the valley with an impetuous 
rush, carrying even stones of a considerable size along with it. 
In other cases landslips had blocked the channel at various 
points for some time, and during the winter snow-slides had 
also aided in forming temporary dams. Even at the end of May 
the stream was spanned at several places by snow bridges. The 
material through which this cutting was made was, of course, very 
soft, and in many places it had been reduced almost to the consistency 
of mud, for all the small tributary streams had been blocked by the 
earth-torrent, and the water was taken up as by a sponge by the 
mixture of loose earth and stones. The photograph reproduced in 
the Plate shows some of the main features of the .erosion but, 
unfortunately, the heavy mists which hung about the hills, lifting 
