6 SURFACE GEOLOGY. 
nearly circular ellipse, so that at one point of its orbit it is somewhat 
farther from the sun than at the opposite point. This eccentricity of the 
earth’s orbit is not constant, but increases and diminishes through long 
periods. During the past fifty thousand years it has been comparatively 
small, and will continue so for the same time to come. The last period 
of great eccentricity began about 240,000 years ago, and lasted 160,000 
years. During this time the winters which occurred farthest from the 
sun, or in aphelion, would be longer and colder than now. The sum- 
mer’s heat would be increased in the same proportion, but it is argued 
that its length would not suffice to melt the annual accumulation of snow. 
This would gain slowly in depth, and become solidified, till a large part of 
this hemisphere would be enveloped in ice. At the same time the oppo- 
site side of the globe would have a short, mild winter, and a long, cool 
summer. Owing to other astronomical causes, known as the precession 
of the equinoxes and motion of the line of apsides, these different climates 
would not be permanent for each hemisphere during the whole of this 
long period, but they would be several times changed, prevailing by turns 
on each side of the equator. In 21,000 years the hemisphere which at 
first had its winter at aphelion would have passed through a cycle, in 
which its place in winter would have traversed the entire orbit,—falling 
after half this time at perihelion, and finally arriving at its first position. 
This theory accordingly supposes that an ice-sheet was produced several 
.times about each pole, alternating with long intervals of genial tempera- 
ture, in which it disappeared. Stratified deposits of sand or clay contain- 
ing organic remains have been found in Europe, underlaid and overlaid 
by till, proving the existence of mild inter-glacial epochs. Equally certain 
proofs of these are rarely found in America. Thick beds of modified 
drift in the midst of till occur in New Hampshire, but they do not appear 
to prove a disappearance and return of the ice-sheet. 
If glacial epochs are produced by a great eccentricity of the earth’s 
orbit, we should also expect indications of ice-action in the older rocks, 
and probably many coarse conglomerates have been formed in this way. 
The remote date to which this theory assigns the last glacial period is not 
improbable, as the amount of erosion effected by Niagara river since the 
ice age, and other facts bearing on this question, indicate a similar lapse 
of time. This, however, seems but as yesterday when it is compared 
