794 



SCIENCE. 



[X. S. Vol. XVIII. Xo. 408. 



can take still another step, but with less con- 

 fidence, in the same direction. 



In reference to the proximate source of the 

 water which changed to steam, which is a 

 conspicuous accompaniment of all volcanic 

 eruptions, there are two leading hypotheses; 

 one, to the effect that it is derived mainly 

 from a deep source and was present in a state 

 of solid solution in the magmas supplied to 

 volcanoes before their migration outward from 

 the earth's interior, and, also, that the steam 

 thus occluded, or its component gases, is ex- 

 pelled as loss of heat occurs and the mag'mas 

 change from a liquid or a plastic to a solid 

 condition; and the other, that water descends 

 from the earth's surface and meets the ascend- 

 ing magmas and becomes occluded in them 

 and at the same time decreases their tem- 

 perature. 



To be sure, both of these hypotheses may 

 be true, and both of the processes referred to 

 be in action at the same time; but even if 

 such were the case, it is to be presumed that 

 one source of water supply or the other would 

 be dominant and in control. 



One of the most interesting questions in 

 connection with the obelisk of Pele is in refer- 

 ence to the evidence it furnishes favoring one 

 or the other of the hypotheses referred to. 



If the steam given ofE by the volcano was 

 an original or primary constituent of the 

 magma which rose in its conduit, it is reason- 

 able to suppose that the distribution of vapor 

 or gases in the magma before its upward 

 migration would be essentially uniform in all 

 parts of the ' reservoir ' from which it was 

 supplied. This assumption, as must be freely 

 admitted, is not susceptible of direct proof, 

 but the little which is known concerning the 

 diffusion of gases in liquids seems to demand 

 that the tension in all parts of a relatively 

 restricted mass of the magma in the subcrustal 

 portion of the earth shall be the same. It 

 may be argued that such a conclusion is not 

 permissible in view of our almost total igno- 

 rance of the condition of matter under the in- 

 fluence of pressures and temperatures such as 

 exist at a depth in the earth, but as the vol- 

 canic problem now stands, it certainly does 

 not seem reasonable to suppose that there can 



be any conspicuous variations from place to 

 place in the primary vapor content of the 

 magma which supplies a single volcano. That 

 is, we have no reason for concluding that the 

 material which was forced upward in the con- 

 duit of Monte Pele, while yet deep in the 

 earth, was in a pronounced degree vapor- 

 charged in one part more than another, and 

 can not appeal to such a supposed variation 

 to account for the diminution in the energy 

 of the explosions in the summit portion of 

 the conduit, or the accompanying change in 

 the material extruded from a fragmental to 

 a massive-solid condition. 



Under the hypothesis that the steam given 

 off by volcanoes has its chief source in the 

 water supplied by downward percolation, or 

 descends from the earth's surface through fis- 

 sures, etc., and meets an ascending magma, 

 the rate of such supply may reasonably be 

 considered as variable and its depletion pos- 

 sible in case great demands are made upon it. 

 How surface water is enabled to reach a 

 volcano's conduit, and the methods by which 

 it is absorbed or passes into a state of solid 

 solution, are again obscure, but these ques- 

 tions may well be left in abeyance, during 

 the search for evidence as to the source or 

 sources from which the water is derived. 



In the instance before us, the evidence 

 seems to show that the earlier explosions 

 exhausted, or at least greatly depleted, the 

 water within reach of the volcano's conduit, 

 and in consequence the conspicuously violent 

 eruptions ceased and the rigid although still 

 hot plug of lava in its summit poi-tion was 

 forced upward by pressure beneath its base and 

 rose far above the rim of its encircling crater. 

 In this connection also it may be noted that 

 the access of water to the summit portion of 

 the conduit of the volcano seems essential to 

 account for the observed cooling and harden- 

 ing of the rising magma at that locality; on 

 the other hand, the fact that the rigid lava 

 was not shattered and blown to fragments 

 suggests that water in considerable quantity 

 did not gain access to it. In reference to 

 these and many other questions bearing on 

 the theory of volcanoes, the reports of the 

 detailed observations that are being made on 



