GLACIAL DRIFT. 235 
they occur, though held for a time in one and the same position while these straight 
lines were engraved upon their surface, nevertheless changed that position more or less 
frequently. A few flatter fragments with more angular outlines show only one kind of 
scratches, having evidently been held for a longer time in the same position. This 
drift, however it may vary in its mineralogical components in different localities, ex- 
hibits everywhere the same characteristic treatment over the whole country, from the 
shores of the Atlantic to the Rocky Mountains, and beyond. In the White Mountain 
region it has the same mineralogical character north and south of the range, and rests 
everywhere upon the well known rvoches moutonnées,—in one word, upon the planed, 
grooved, polished, and scratched surfaces of the rocks underlying it. 
Observation has taught us that materials such as those described above, so com- 
bined, exhibiting the same characters in their surfaces, and having the same diversity 
of composition and absence of all sorting or regular arrangement, occur now at the 
bottom of the great glaciers of our time, and nowhere else, being found between the 
ice and the rocks over which it moves,—the result, in fact, of the grinding action of ad- 
vancing glaciers. On account of their unvarying position, I have called these deposits 
‘‘ground moraines,” because they are always resting upon the rocky floor of the coun- 
try, between it and the under surface of the ice. Our typical unaltered so-called 
northern drift is synonymous with the ground moraines of the present day, differing 
only in its greater extension. It is in facta ground moraine spreading over the great- 
est part of the continent. All its characteristics, identical in every detail with those of 
the deposits underlying the present glaciers, show that it can only have been formed 
under a moving body of ice, held between it and the underlying mass of rock. The 
great ice sheet of the glacial period which fashioned the drift must therefore have been 
coéxtensive with the distribution of the latter. It is very important to distinguish this 
drift from the moraines formed under other circumstances, and from the so-called 
erratics and perched blocks. Moraines, as commonly understood, that is, lateral and 
frontal moraines, consisting of loose materials collected along the sides and at the 
terminus of a glacier, always indicate, and, where undisturbed, actually define the mar- 
gins of a moving mass of ice; whereas, the so-called median moraines formed along the 
line of junction of the glaciers are carried upon the back or upper surface of the ice, 
and always consist of angular materials, the shape and arrangement of which are de- 
termined by their mode of accumulation. Just as among the glaciers of the present 
day we discriminate between ground moraines, lateral, frontal, and median moraines, 
so must we also distinguish between the same phenomena in past times. The glacial 
period had also its ground moraines, its lateral, its frontal, and its median moraines, 
its erratics and perched boulders. But the huge ground moraine of the earlier ice 
time stretched continuously, like the ice sheet under which it was formed, over the 
whole country, from the arctics to the Southern states, and from the Atlantic to the 
Rocky Mountains. I do not speak of the western slope of the continent, because I 
have not examined it personally. The great angular erratics of that period were scat- 
tered irregularly over the country, as the few large boulders are scattered on the upper 
