other reasons than those given by Mr. Croll why its effects may be greatly varied. Nor can there be any doubt that changes in the general level of the land, altering coast-lines and the currents of both the atmosphere and the ocean, are important agents in modifying temperatures. We cannot follow into details Mr. Croll's exhaustive inquiries, but our readers will be interested in his answer to the well-known theory of Sir Charles Lyell. According to that eminent authority, a period of polar cold will result from a great increase of elevation and extent of land in polar regions, and a warm period in the polar zones will occur by a great accession of land in the equatorial regions. But Mr. Croll shows that such changes in the distribution of land would be followed by opposite results—that a great accession of land in the equatorial zone would destroy the system of ocean-circulation by which the heat of the equator is made to do service in warming the ocean and the air of colder zones.
In Mr. Croll's theory it is impossible that both hemispheres should be glaciated at the same time. Not only must periods of heat alternate with those of cold in one hemisphere, but a glacial epoch in one is accompanied by a temperate epoch in the other.
It is evident that changes of climate such as are shown to have occurred must have arisen from general, not from any local cause or accidental combination of causes, and if this be so there may yet appear a reliable means of determining not only the amount and extent of the changes, but the periods of their duration and recurrence. The question of time in geological history is an important and certainly a most interesting one, and the interest in it has increased since the announcement by Dr. Tiddeman and others that human relics have been found in deposits of the warm inter-glacial periods.
Two methods have been adopted by Mr. Croll which are supposed to throw light on this subject. One is to fix the period and duration of the epochs of greatest heat and cold by computing the period and duration of the astronomical coincidences already noticed, by which according to his theory those epochs were brought about. The other method is applied to estimating the time since the close of the last glacial epoch by changes known to have taken place in the earth's surface, and the general lowering of the land by denudation. To do this, he says, "we have only to ascertain the quantity of sediment annually carried down by the river systems."
By this means it is found that the lowering may have been nearly a foot in 6,000 years. But when we consider how greatly the general result may have been interfered with by the alternate elevation and depression of the land, a work now going on, we realize that Mr. Croll's conclusions do little more than profoundly impress the mind with the vastness of time required in some of the most obvious of Nature's operations.
By the first-named method computations have been made extending
