THE ORIGIN OF GUMBOTIL 119 



loess and loesslike clay are of great interest but are not being 

 considered except incidentally in this paper. In places these 

 materials are clearly eolian in origin. In other places the loesslike 

 clay may be related closely in origin to the gumbotil. The presence 

 of loess or loesslike clay above the gumbotil might be expected to 

 have had some slight effect upon the present chemical composition 

 of the gumbotil, and in fact may explain the few percentages in the 

 analyses which might seem to contradict the theory proposed. 



It is assumed that the composition of the flour of the unoxidized 

 and unleached till is now the same as it was when laid down by 

 the glacier. The possibility exists, however, that it may have 

 received a small amount of leached material from above, or that it 

 may have lost to the strata above by capillary flow slight quantities 

 of the more easily diffusible dissolved materials. It is not to be 

 expected that its composition will be similar to that of overlying 

 materials which have been subjected to marked chemical changes, 

 to leaching, or to infiltration. 



Attention should be called again to the fact that the gumbotils 

 occupy definite topographic positions and that, as a result of 

 erosion subsequent to the formation of the respective gumbotils, the 

 areas of gumbotil are now very limited compared with the extent 

 of the former gumbotil plains. 



A study of Tables I to IV will bring out several interesting 

 facts. In all of the series here represented the percentage of AI2O3 

 decreases downward from the gumbotil through the oxidized and 

 leached zone. Except in the case of "B " (Table II), this decrease 

 continues also through the oxidized and unleached zone. 



Perhaps the most important evidence in favor of the leaching 

 theory is to be gained from a study of the relative proportions of 

 CaO and MgO in the various horizons. In practically every series 

 the proportions of these two constituents show a pronounced 

 increase downward. Apparent contradictions for both might be 

 considered for MgO in Table III and CaO in Table IV. Field 

 relations will show in these instances either that the gumbotil is 

 overlaid by material containing a higher proportion of these con- 

 stituents, or that erosion began before the leaching process in the 

 gumbotil was completed. Assuming that the loess or the loesslike 



