514 THE JOURNAL OF GEOLOGY. 
approximately at least—to points under consideration for inter- 
preting properly the changes as shown by the photographs. 
The glacier having been plotted as already stated, any required 
distance was obtainable. 
A question that suggested itself from the photographs was 
that of determining the shrinkagé or melting of the ice (abla- 
tion) between July 13 and August 11, irrespective of any 
motion. This was obtained in the following manner: On the 
photographs well-marked points were selected on the adjacent 
mountains, verticals through which cut the glacier. Each point 
selected was common to both views. Evidently points which lie 
in the same vertical plane in one photograph lie in the same 
plane in another photograph from the same station. There were 
thus eighteen intersections common to two photographs of the 
crest of the face of the glacier obtained. We hence had the 
angular measurement of these intersections, and as the distances 
thereto were sufficiently well known, the linear measure followed. 
It is not forgotten that, especially near the face of the glacier, 
the motion of the ice causes crevasses and upheavals so that in 
a particular vertical section there may be an abnornal decrease 
or increase in elevation and the phenomenon of shrinkage obliter- 
ated or at least hidden. In fact, the observations or photographs 
show such to be the case; for in one instance we have a rise of 
nearly six feet, instead of a decrease of two feet as the average 
shows. From these eighteen intersections we find that during 
the twenty-nine days,—July 13 to August 11,—the mass of the 
glacier—ice front—fell a little over two feet (2.1 feet). 
Let us now examine the linear motion of the ice. This 
problem we can attack from two points — using only photographs 
from one (nearest— No. 3) station.— Firstly, by finding the ver- 
tical motion, and secondly by determining the change in azimuth. 
The former is directly obtained by measuring on the photographs 
the distance of a point on the two photographs, above the hori- 
zon line, then from its known distance the absolute height in 
each case is obtained, and the difference will be the motion in 
altitude for the interval of the photographs. From the general 
