THE PROXIMATE CAUSE OF ERUPTIONS. 127 



venient season to explode. It presents the case as a problem of energy 

 acquired by some secondary forces, of which we are at present ignorant. 



There is one general assumption which satisfies all the main requisites 

 of volcanism. It is this : Volcanic phenomena are brought about by a local 

 increase of temperature within certain subterranean horizons. This, indeed, is 

 not a solution of the problem, for it throws us back instantly upon the ulte- 

 rior question, What has caused the increase of temperature 1 All my efforts 

 to find an answer to this ulterior question have utterly failed. But the 

 proximate idea is suggested on every hand, and its reality takes deeper root 

 in conviction the more it is contemplated. Around it the broader facts take 

 form and coherence. It explains their secondary character as contradis- 

 tinguished from the primordial. It explains the cyclical phases of volcan- 

 ism; their beginning in a recent epoch of the world's secular history; their 

 growth, decay, and extinction. It explains their intermittent character — 

 why eruptions are repetitive instead of continuous. It explains the explo- 

 sive and energetic character of the phenomena ; and, lastly, it explains the 

 lithological order of the eruptions, as will presently be shown. 



But there is another and alternative assumption. We may suppose 

 the deeply-seated rocks in regions of high temperature to undergo changes, 

 one result of which is to lower their melting-points. Tins is not so strange 

 as it might at first seem, for its accomplishment is conceivably within 

 known physical laws. A relief of pressure is one conceivable mode. 

 Probably another would be the absorption of water under great pressure 

 and at high temperature. It can hardly be doubted that a rock charged 

 with water and so confined that the water cannot readily escape is more 

 fusible than the same rock in an anhydrous condition. The fact that 

 lavas bring to the surface considerable quantities of water may be held to 

 be evidence that water does find access to them from above. The only 

 alternative view is that water formed a part of their original constitution. 

 This is undoubtedly the case on the view that lavas are remelted metamor- 

 phic rocks ; for the metamorphics all contain water, partly mechanically 

 held and partly as water of combination in hydrous minerals. The amount 

 of contained water is variable, but ordinarily more than one per cent, and 

 sometimes much more This quantity, however, probably falls far below 



