the gases projected by the sun beyond its gravitative control, in interstellar spaces, in meteorites and in the terrestrial mass. The first three possible sources are too indefinite both in amount and in distribution through the ages to be of any present value to the hypothesis, but the last is important. Crystalline rocks of the superficial crust of the earth are shown by analysis to contain four and one-half times their own volume of gases, of which carbon dioxide, CO2, and monoxide, CO, form a large percentage. During volcanic eruptions gases and vapors are ejected in indefinitely large volumes. What part of these was once of the atmosphere and is returned to it after an underground journey we do not know, but it is believed that a large part may come from the interior. The escape of these internal gases may also occur in some degree continuously by diffusion, and in influential amounts during episodes of mountain growth, when rock masses are strained and riven and upraised. We shall see presently that in the wasting of a mountain range there is serious consumption of carbon dioxide, which in greater or less degree temporarily affects the gain. The Pleistocene glaciation is attributed to a very notable offset of this character, but the exceptional nature of that event as compared with the relatively frequent episodes of mountain growth indicates that the gain of carbon dioxide has commonly equaled or exceeded the resulting temporary loss.
Among the cycles of combination and release, through which carbon dioxide runs, there are two which are both important, though not equally so. The first and less important is the cycle of organic change, involving plant and animal tissues. When grass grows, carbon dioxide is taken from the air. Grass becomes beef, and beef, through, various changes', is resolved into new compounds, yielding back the carbon dioxide to the air. In its brief phases this cycle has no import for the hypothesis, but there are occasions where it is prolonged, as in the accumulations of vegetal substances fossilized as coal. The total amount of carbon dioxide thus abstracted, and now withheld, is very large, and is believed to have been an important factor in promoting at least one instance of glaciation, that which followed closely upon the Coal Measure period.
The more important cycle of combination and release involves the decomposition of rocks by weathering, the solution of certain products and their transportation to the sea, and the reactions through physical, chemical and organic agencies, by which carbon dioxide is either permanently locked up in limestones or is returned to the air.
For the purposes of this statement, the common minerals of rocks may be classified as silicates and carbonates, that is as compounds with silicic acid and compounds with carbonic acid. The former may be typified by the familiar minerals of granite, the latter by limestone.
