266 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



on the day followin<^ the explosion noted two cracks extending east 

 and Avest from the explosion opening for a hundred feet or more. 

 The explosion itself, so far as may be gathered from witnesses, was 

 wholly inadequate in volume to split open the mountain, and indeed 

 the small crater formed by the explosion is reported to have been no 

 more than 20 by 40 feet and extended only a few feet into the loose 

 scoriaceous material covering the crater floor. 



This discharge of steam in violent explosion was followed by 

 similar outbreaks at intervals of three or four days during the next 

 seven or eight months, the periodicity being determined no doubt by 

 the rate of crystallization and the amount of the overlying load. In 

 general the volume and intensity of the successive explosions increased 

 as they proceeded, which might reasonably be expected, as more 

 and more of the ciystallizing mass came to feel the release of pres- 

 sure after the opening had been made. 



The laboratory study of such solutions indicates a further possi- 

 bility through which the early stages of such an outbreak might be 

 expedited. The fact that even when the explosions had rea(;hed their 

 maximum intensity there was no development of high temperature, 

 suggests that the magma below may itself have been at a relatively 

 low temperature, ultraviscous and inert through undercooling, that 

 is to say, being inclosed within an impervious containing vessel at 

 comparatively low ti mperature, crystallization had slowed down 

 until considerable portions were cooled below the temperature of 

 normal crystallization while still in the vitreous state. It is of course 

 well known that silicate solutions (cf. window glass) are prone to 

 crystallize with extreme slowness, otherwise probably we should 

 have no glass industry. In such an inert mass therefore in which 

 crystallization was proceeding thus slowly, the advent of water vapor 

 from without would have the effect of diminishing the viscosity and 

 increasing the rate of crystallization. 



Following the first explosion at the summit of Lassen Peak it was 

 noted that the great body of accumulated snow within the summit 

 crater was rapidly melting and flowing into the explosion cracks. 

 This water must have found its way to the volcano hearth at first 

 by percolation and later, within the hot zon ■, as steam. We know 

 from laboratory experiment that steam is readily soluble in such 

 an undercooled silicate magma with the effect of increasing its 

 fluidity and thereby precipitating crystallization or increasing the 

 rate of crystallization already going on. The operation of crys- 

 tallization itself, as has been shown above, sets free more water so 

 1 hat the operation becomes more or less self-p;'rpetuating so long as 

 any considerable amount of magma remains uncrystallized within 

 reach of the outlet channels (reduced pressure). This explanation 



