572 THE JOURNAL OF GEOLOGY. 
to be confined to very definite planes. In some places there was 
a somewhat promiscuous scattering of the rock material through 
the ice, and in exceptional cases considerable thicknesses of ice 
were freely inset with foreign matter, but as a rule the erratic 
material was distributed in extremely thin sheets between layers 
of ice that remained essentially pure. The typical section was 
therefore made up of layers of pure, clean ice separated by films 
of rocky débris. The rocky material varied in size from the 
finest silt up through pebbles and fragments of various sizes to 
bowlders or blocks of rock several feet in diameter. There was 
no special assortment of this material. Between the same two 
layers of ice might be found, here an attenuated film of fine silt, 
Se ea SS 
SS SSS SSO 
——as = SaaS —— 
———S SSS ——— 
a a SSeS 
Fic. 37.—Diagrams illustrating the behavior of the laminze of the ice in passing 
embedded bowlders. 
a little farther on a layer of sand, or a pebble, or a chip of rock, 
or all these together, while now and then a bowlder or massive 
block might be encountered. It was observed that usually these 
larger pieces centered upon the plane of débris, a portion of their 
mass rising above it and a portion sinking below it. Where the 
ice was closely laminated the larger fragments necessarily 
extended across the horizon of several lamine, and it was inter- 
esting to observe that in many of these cases the lamine divided, 
a part bending up and passing over the rock fragment and a part 
bending down and passing under. In cases where the mass was 
large, some of the central laminations ended abruptly against the 
mass and new laminations appeared on the opposite side, while 
the laminz above and below were bent around the mass as illus- 
trated in Fig. 37. I observed no cases in which the mass seemed 
to have descended through the ice, as though carried down by 
its superior gravity. In sucha case it would be presumed to 
bend the lamin down with it, or break across them. It appeared, 
