6o4 CLARENCE N. FENNER 



magma, and the question may be asked as to what combinations are 

 supposed to be entered into between the volatile and non-volatile 

 constitutents that would give the required effects, also as to the 

 nature of the changes in, the physical environment by which the 

 condition of unstable equilibrium is brought about. These are 

 natural queries, and answers would be eminently desirable; never- 

 theless, it is believed that the case rests not upon the ability to 

 answer them but rather upon the plain evidence of the phenomena 

 themselves. It may be of assistance, however, to a comprehension 

 of what is meant by unstable equilibrium of the magma, to con- 

 sider certain familiar phenomena exhibited by obsidians. Any 

 obsidian which, when heated, puffs up into pumice, shows charac- 

 teristics allied to those that I have ascribed to the Katmai magma. 

 The behavior of the obsidian in this respect indicates that it like- 

 wise in its past history underwent changes of condition in which 

 internal equilibrium failed to keep up with external changes. If 

 this were not true it could hardly have retained its dissolved gases 

 but would have evolved them during cooling. In instances of 

 this kind the lack of equilibrium continued even beyond the 

 stage which the Katmai lavas reached, and finally all possibility 

 of evolving gases disappeared because of increasing rigidity, but 

 this result was probably due to factors (such as rate of cooling) 

 which may well be variable. In the case of Katmai and other 

 volcanoes, it seems reasonable to suppose that the magma first 

 experienced very rapid change of conditions during its rise in the 

 conduit, but then remained for a certain period in comparative 

 quiescence, and thus opportunity was given for approximate 

 equilibrium to be reached before a condition of prohibitive rigidity 

 had set in.' 



From evidence of the kind given it appears that examples of 

 unstable equilibrium in magmas, due to sudden changes of environ- 

 ment, are not at all uncommon. Recognition of this fact and of 

 what it connotes may be helpful in directing inquiry into the con- 

 ditions that have brought it about. 



^ Interesting examples of the effect of rate of cooling upon the final product in 

 somewhat analogous systems are furnished by Moray's experiments on hydrated 

 alkali silicate melts prepared in steel bombs. Rapid quenching gave a rigid, hydrated, 

 unstable glass, while slow cooling caused the expulsion of dissolved water and the 

 formation of a pumice. See G. W. Morey, Jour. Amer. Chem. Soc, Vol. XXXVI 

 (February, 1914), p. 226. 



