264 EEPORT UNITED STATES GEOLOGICAL SURVEY. 



largest portion of the lava. Carbonates, sulpliurets, and other combined 

 forms occur comparatively sparingly, and must be regarded, in many 

 instances, as epigene products. 



From the evidence adduced by erupted material, therefore, we can 

 infer that those portions of the earth's crust which furnish the ejected 

 masses are composed of direct oxygen compounds and silicates mainly. 

 These two classes make up by far the largest i^roportion of the crust, so far 

 as it has been possible to study it. Several of the elements must, there- 

 fore, be accounted for: hydrogen, nitrogen, chlorine, carbon, and sul- 

 phur. In attempting to explain the presence of these elements at very 

 considerable depths, we touch upon a subject which, in this connection, 

 becomes one of the utmost importance. This is the subject of infiltra- 

 tion. 



The emission of water during volcanic eruptions has been a subject of 

 careful study. From the fact that active volcanoes are found almost 

 exclusively in the vicinity of large bodies of water, the hypothesis has 

 been derived that they are in direct connection with such bodies. A 

 number of theories have been advanced explanatory of the method 

 whereby the water could find access to the regions of high temperature. 

 Others have assumed that the aqueous vapors emitted were merely the 

 equivalent of precipitated moisture or of direct leakage. Either one of 

 the views undoubtedly expresses a portion of the truth. It is eminently 

 probable that water precipitated in the immediate vicinity of volcanic 

 vents may find its way, as drainage, to such depths, from which it will 

 be ejected by force. On the other hand, fissures, in connection, perhaps, 

 with standing bodies of water, may conduct a certain amount of it to 

 the upper portions of the liquefied rock. Besides these sources, how- 

 ever, we have that of the perpetual percolation of water to great depths. 

 It is an accepted theory that water penetrates the rocks composing the 

 earth's crust to a far greater depth than has been reached by man. We 

 have here, then, a prolific source of oxygen and hydrogen. Carbonic- 

 acid gas is equally assumed to penetrate to very great depths. As a 

 concomitant of water it is well known. Sulphates and chlorides can be 

 found in aqueous solutions, and chlorides are the main mineral constitu- 

 ents of sea-water, readily held in solution. Sir H. Davy and others have 

 assumed the jfresence of atmospheric air in volcanoes, occurring at depths 

 sufficiently great for the purpose of decomposition. We have, there- 

 fore, in the afflux of water and the compounds it usually holds, either in 

 solution or mechanical retention, the means of explaining the presence 

 of all gaseous elements or compounds observed in connection with vol- 

 canic eruptions. 



Having thus briefly alluded to the evidence furnished by volcanic 

 eruptions referable to admixtures that can be regarded as accidental, 

 when compared with the lava itself, we proceed to the consideration of 

 the causes producing such eruptions. 



In the presence of water at depths where the temperature must show 

 a very high degree, we have a factor that is capable of very great ex- 

 pansive power. Analogous to eruptions of geysers, the eruptions of 

 volcanoes have frequently been regarded as being due to the change of 

 water into steam. No more powerful agent, perhaps, could be imagined 

 than steam for propelling through existing fissures the liquefied rock — 

 the lava. Accepting the theory which seeks the origin of lava not in 

 the central fluid or plastic masses of the earth, but in "reservoirs," more 

 or less localized, it becomes incumbent to account for the liquefaction of 

 those portions of rigid rock which have filled the reservoirs. 



l^Tumerous views, intended to explain the formation of such reservoirs 



