354 The Glacial Theory of Prof. Agassiz. 
» The ascent of blocks from the middle or lower part of the ice to the 
surface, explains another curious fact—that though the general motion 
of the glacier is along an inclined plane downwards, scratches are ofien 
found on the rock inclined in the opposite direction ; that is to say, 
supposing the surface of the glacier to dip at 10 degrees to the north, 
you will find scratches dipping at 10 or 20 degrees to the south, or 
even vertical. ‘These are caused, in Agassiz’s opinion, sometimes by 
inequalities in the bottom of the valley, but frequently by enclosed 
blocks working their way upward by the expansion of the ice, while 
the glacier is travelling downwards. 
Figure 2 represents the usual form wi a lateral moraine in the cross 
section, and as it would appear on a surface considerably inclined ; m 
the mass of gravel form- | Fig. 2. one > Fig. 3. 
raine would have if the ice were melted, and the matter left on the 
surface of the valley. 
Retreating glaciers form a terminal moraine every year, as mention- 
ed in page 353; and in 
this case we might expect 
to find a series of mounds. | 
transverse to the valley =. 
like x y, figure 5. 
. Erratic Blocks.—Sin- |= 
$e blocks of huge size — 
are ofien seen resting on the surface of the glacier, and ioesili 
downwards with it. ‘These are generally angular, and they often stand 
‘on pedestals of ice, as in figure 4, where a is a tabular mass of rock, 
and’d the pedestal of ice.* Agassiz describes one he saw on a glacier, 
which measured 20 feet by 12, and must have weighed 100 tons or 
more. «In accounting for the pedestal b, he observes that gravel, when 
it 'rests on the surface of a glacier, being heated through and through 
wy the-sun’s oe melts the ice below it, and gradually forms a pool or 
well init A. “A latge block, on the other hand, has only its upper sur- 
whilethe*inferior mass, remaining cold, protects the ice 
below=both fromi the: action of the sun’s rays, and from the evapora- 
ee like water, “oe ill in the open 2 airy mee, 
Fig. 4. Fig. 5. 
