THE KATMAI REGION, ALASKA 599 



among the grains may have had a melting-range Httle above the 

 temperature of the new magma, and the penetration of the latter 

 and of its vapors was therefore easy; second, an original vesicu- 

 larity may have been present; and third, a porous condition may 

 have been developed during the history of the rock. The third 

 feature is important. It is not hypothetical, but rests upon 

 observation, and is believed to have considerable significance. It 

 has been found that many of the andesitic fragments in the pumice- 

 ous strata have at some former time undergone a process of altera- 

 tion similar to that described for the rocks of Falling Mountain. 

 Quantities of minute, glistening scales of tridymite have been 

 formed, and a replacement of ferromagnesian minerals by hema- 

 tite is observable. This carries several implications: first, these 

 mineral transformations are such as might be expected to result 

 from fumarolic action along fissures in the walls of a crater, but 

 not of the kind that would be looked for at great depths; second, 

 the presence of such fissures and the mineral transformations 

 along them would aid in the collapse of the walls during the activities 

 of the eruption; and third, these altered rocks would be more 

 susceptible to penetration by the magma on their immersion in it. 

 When the process of penetration of magma into porous xenoliths 

 has been thorough, they appear to have become pasty throughout, 

 and what we find is an irregular, lumpy mass, or clot, consisting of 

 basic minerals in a dark groundmass. Under the microscope the 

 phenocrysts are seen to be much corroded. They often contain a 

 great number of inclusions of brown glass, which fairly riddle them, 

 and they look as if they were disintegrating. The groundmass is 

 essentially glassy but contains a multitude of small, irregular frag- 

 ments or splinters of crystals, and much brown dust. The low 

 index of this glass indicates an acid material in spite of the dark 

 color, and it seems doubtful whether the amount of material 

 actually fused or dissolved was large; the process was rather one 

 of intimate penetration by the new magma, resulting in separation 

 and dispersal of the component minerals and a partial breaking 

 up of crystal units. The low silica-content found by analysis (as 

 in the specimen illustrated in Fig. 11) and the dark color are 

 probably due to undissolved phenocrysts and dust. When the 



