POPOCATEPETL AND IXTACCIHUA1L—FARRINGTON. 113 
glacier of the second order, the terminal moraine of which is the 
deposit which I have mentioned, instead of being regarded as trib- 
utary to the main glacier. At times, however, without doubt, this 
portion of the glacier joins the main body. 
MIDDLE COURSE OF THE GLACIER. Though lateral moraines 
are wanting, there is no lack of evidence that the glacier is a true ice 
stream constantly bearing and depositing debris. Scattered promis- 
cuously over the surface of the terminal portion of the glacier lie 
bowlders of all sizes, as well as other varieties of rock debris which 
could have reached their positions only by being transported by the 
ice. 
The fact is illustrated in Pl. XVI, a view taken on the surface of 
the glacier several hundred yards from its terminus and about 200 
feet south of the northern border. The point at which the photograph 
was taken is indicated ata, Pl. XV. The great bowlders here shown 
must have floated to their position on the ice stream. Other features 
common to the middle course of glaciers could well be seen in this 
vicinity. There were glacial tables formed by bowlders sometimes 
several cubic feet in dimension, resting on pillars of ice 12 to 18 
inches in height; sand cones 2 to 5 feet in height and ridges similarly 
formed 20 or 30 feet in length. As is well known, and as the differ- 
ent stages of formation well showed, these various phenomena have 
their origin in the protection from melting furnished to the ice by the 
sand or bowlders above it, which brings the latter into relief as the 
adjacent ice melts away. 
The surface of the ice was nowhere strongly crevassed. The 
crevasses which existed were irregular and short, few being over 
30 feetin length. All that were observed had a course transverse 
to that of the glacier, and thus, as might be expected, no well marked 
séracs were formed. The fantastic, pinnacled shapes assumed by the 
upper surface of the ice, illustrated in Pl. XVI, indicate, however, a 
certain amount of longitudinal fracturing. 
It was of interest to note the gathering of surface bowlders by 
the crevasses causing them to lie in rows across the field of ice. This 
formation was produced by the more rapid melting of one wall of the 
crevasse than the other, forming a long slope, which dipped toward 
the other nearly perpendicular wall. Along this the freed bowlders 
slid to the crevasse as illustrated in Fig. 2, and so were gathered in 
rows following its direction. Thesloping side was usually the eastern. 
Iam of the opinion that some of the minor bowlder trains found 
lying transverse to the direction of glacial movement in the glaciated 
regions of our own country may have had a similar origin, since the 
