THE KATMAI REGION, ALASKA 595 



cliff-walls to support themselves, but this may be quite independent 

 of a general subsidence of the floor of the crater. We know from 

 Martin's account^ that natives who had apparently fled from the 

 region almost at the beginning of the activities reported that the 

 top of the mountain was gone. We should hardly expect crater 

 subsidence to take place at this early stage. On the whole, it 

 seems that this idea should be applied only if the process of solu- 

 tion, which, in any case, seems to have occurred on a large scale, 

 appears quantitatively inadequate to account for all the material 

 that has disappeared. At present it seems best to postpone 

 judgment on this until more information is obtained from the 

 analyses. 



EVIDENCE AS TO THE NATURE OF THE ERUPTIVE PROCESSES 



A study of the ejecta from Katmai and of the characteristics 

 of the deposits that they form supplies considerable additional 

 information on the eruptive processes. When seen in undisturbed 

 deposits at a distance of, say, eight or ten miles from the crater, 

 the ejected matter forms well-defined strata, such as are shown in 

 Figure 17. The component material is chiefly a light-gray pumice 

 in pieces whose dimensions are three to four inches as an ordinary 

 maximum, and run from this to a very minute size. Mingled 

 with this are specimens of banded pumice, dense obsidian, stony 

 andesites, sedimentary shales, and a sort of volcanic conglomerate. 

 Some of the features of these, and their significance, have been 

 touched upon before but will now be considered in more detail. 



It was pointed out, in discussing the great sand-flow in the 

 Valley of Ten Thousand Smokes, that the banded and variegated 

 character of some of the pumice indicated a mixture of basic 

 material with the new siliceous magma shortly before extrusion. 

 In such specimens, sharply defined bands adjacent to each other 

 show such dift'erences of composition as are indicated by Dr. Allen's 

 determination of 74 . 70 per cent silica in one and 60 . 40 per cent 

 silica in another. The material thrown out from Katmai crater con- 

 tains similar specimens. The dark bands consist of partly digested 

 basic rock with large quantities of minerals appropriate to andesites. 



' G. C. Martin, Nat. Geog. Mag., Vol. XXIV (February, 1913), No. 2, p. 147. 



