The shifting of the beds by faults is often intimately connected with those same foldings which constitute the anticlinal and synclinal axes before alluded to, and there is no doubt that the subterranean causes of both forms of disturbance are to a great extent the same. A fault in Virginia, believed to imply a displacement of several thousand feet, has been traced for more than eighty miles in the same direction as the foldings of the Appalachian chain.[13] An hypothesis which attributes such a change of position to a succession of movements, is far preferable to any theory which assumes each fault to have been accomplished by a single upcast or downthrow of several thousand feet. For we know that there are operations now in progress, at great depths in the interior of the earth, by which both large and small tracts of ground are made to rise above and sink below their former level, some slowly and insensibly, others suddenly and by starts, a few feet or yards at a time; whereas there are no grounds for believing that, during the last 3000 years at least, any regions have been either upheaved or depressed, at a single stroke, to the amount of several hundred, much less several thousand feet. It is certainly not easy to understand how in the subterranean regions one mass of solid rock should have been folded up by a continued series of movements, while another mass in contact, or only separated by a line of fissure, has remained stationary or has perhaps subsided. But every volcano, by the intermittent action of the steam, gases, and lava evolved during an eruption, helps us to form some idea of the manner in which such operations take place. For eruptions are repeated at uncertain intervals throughout the whole or a large part of a geological period, some of the surrounding and contiguous districts remaining quite undisturbed. And in most of the instances with which we are best acquainted the emission of lava, scoria, and steam is accompanied by the uplifting of the solid crust. Thus in Vesuvius, Etna, the Madeiras, the Canary Islands, and the Azores there is evidence of marine deposits of recent and tertiary date having been elevated to the height of a thousand feet, and sometimes more, since the commencement of the volcanic explosions. There is, moreover, a general tendency in contemporaneous volcanic vents to affect a linear arrangement, extending in some instances, as in the Andes or the Indian Archipelago, to distances equalling half the circumference of the globe. Where volcanic heat, therefore, operates at such a depth as not to obtain vent at the surface, in the form of an eruption, it may nevertheless be conceived to give rise to upheavals, foldings, and faults in certain linear tracts. And marine denudation, to be treated of in the next chapter, will help us to understand why that which should be the protruding portion of the faulted rocks is missing at the surface.