state. This fact serves to show us that the changes necessary to
produce great ice sheets are not necessarily of a startling nature,
however great the consequences may be. Assuming, then, that relatively
slight alterations of climate may cause the ice sheet to come and go,
we may say that all the influences which have been suggested by the
students of glaciation, and various other slighter causes which can
not be here noted, may have co-operated to produce the peculiar
result. In this equation geographic change has affected the course of
the ocean currents, and has probably been the most influential, or at
least the commonest, cause to which we must attribute the extension of
ice sheets. Next, alterations of the solar heat may be looked to as a
change-bringing action; unfortunately, however, we have no direct
evidence that this is an efficient cause. Thirdly, the variations in
the eccentricity of the earth's orbit, combined with the precession of
the equinoxes and the rotation of the apsides, may be regarded as
operative. The last of all, changes in the constitution of the
atmosphere, have to be taken into account. To these must be added, as
before remarked, many less important actions which influence this
marvellously delicate machine, the work of which is expressed in the
phenomena assembled under the name of climate.

Evidence is slowly accumulating which serves to show that glacial
periods of greater or less importance have been of frequent occurrence
at all stages in the history of the earth of which we have a distinct
record. As these accidents write their history upon the ground alone,
and in a way impermanently, it is difficult to trace the ice times of
ancient geological periods. The scratches on the bed rocks, and the
accumulations of detritus formed as the ice disappeared, have alike
been worn away by the agents of decay. Nevertheless, we can trace here
and there in the older strata accumulations of pebbly matter often