in temperature in the rocks above, and thus by expanding them, increase their upward tendency. The subsequent cooling of these heated and possibly metamorphosed rocks would also tend to renewed subsidence of the surface. In other words, subterranean intrusions are accompanied by an abnormal rise of the isogeotherms and their loss of heat by a return to normal conditions.
In progressively loaded areas, as has been pointed out by Reade, Shaler, and others, there is a blanketing of the earth's heat and a rise of the isogeotherms; the accompanying expansion of the rocks tends to check subsidence and limit accumulation. The processes of erosion, transportation, and sedimentation in special areas are thus limited by conditions within the earth. Erosion favors elevation until the plastic material transferred to the region beneath the lightened area cools and hardens. A decrease in elevation due to contraction then ensues, and is accompanied by a decrease in erosion, which comes to an end when the land is reduced to base level. Loading favors sedimentation by causing subsidence until the thickened sediments become heated and by reason of their expansion elevate the surface to or above base level. There is a mutual interaction beneath the earth's crust also, since, if only one region of erosion and one of sedimentation are considered, the checking of elevation in a region of erosion by the cooling and hardening of injected magma beneath will give greater resistance to the flow of plastic material from beneath the region of sedimentation.
Deep erosion of subtuberant mountains should reveal a central area of igneous rock surrounded by a belt of metamorphosed rocks which on its outer border, in case the injection occurred in ordinary sedimentary strata, should pass into unaltered sand-stone, shale, etc. In such an instance a radial section should reveal a gradation from igneous rock through metamorphosed rocks to unaltered sedimentary beds. The breadth of the central core of igneous rock would vary with the size of the intrusion, and, down to a certain limit at least, with the depth of the plane of erosion. One or more generations of dikes might occur in either the central area or in metamorphosed or sedimentary rocks surrounding it. Great intrusions if deeply eroded would thus present the conditions sometimes cited as examples of "regional metamorphism." Some of the features observed in the crystalline region of Canada seem to illustrate the surface features that would be found if a great subtuberant uplift should be pared away by erosion.
The fact that a large majority of volcanoes are situated near the sea has led to the supposition that sea water gaining access to highly heated rocks is the chief if not the essential cause of volcanic eruptions. The hypothesis, however, that the sea is the
