246 N. S. SIIALER — LOKSS DEPOSITS OF MONTANA 



on tlie (letrital cone of Alder creek, which lies in the broad valley of 

 that river. 



Conditions detkumininu Formation of Loess 



The observations which were made justify the general statement that 

 wherever the surface of this prevailingly arid district is moistened by the 

 streams which head in the snow-bearing uplands we find an accumula- 

 tion of loess which in all cases is somewhat, mingled with decayed vege- 

 table matter, the amount of this matter being so variable that the material 

 ranges from a dark loam to an impure soft peat. The thickness of the 

 deposit varies from a few inches to 10 feet or more, being usually greatest 

 in depth where the vegetation covering the surface is richest and where 

 on that account the humus is in largest proportion. Eliminating the 

 organic matter, it will probably be found that the extreme thickness of 

 this deposit, except where it has been washed from its original site and 

 accumulated in low places, does not exceed 6 feet. 



The conditions of the formation of this loess deposit may be observed 

 at any time when the earth is dry and the wind strong enough to lift the 

 dust, as is the case for a considerable part of each year. From the sur- 

 face of the benches of the valleys, as well as from the scantily vegetated 

 lower parts of the mountain ranges, dust is blown to and fro in large 

 quantities. So long as it encounters no closely set vegetation it does not 

 come to rest. It is only when it finds its way amid densely set plants 

 in the limited areas watered by the snow-fed streams that it escapes from 

 the controlling wind. In such places it is quickly fixed to remain so as 

 long as the natural or artificial irrigation continues. 



Effects of Loess on Streams 



The first consequence of these local accumulations of dust is to be seen 

 in their effect on the course of the streams about which they accumulate. 

 The fine detritus makes a very rich soil, so that a luxuriant vegetation is 

 nurtured. This tangle is likel}' to obstruct the movements of the water, 

 as the detritus brought down in times of Hood lacks the chance of lateral 

 escape which is commonly afforded on detrital cones where loess is not 

 thus deposited. The result is that the bed of the rivulet soon becomes 

 so far raised that it is forced to break a new path, but so long as the water 

 is confined bv the dense wall of vegetation on either side it is forced to 

 convey its burden farther out into the valle}^ than would be the case on 

 ordinary barren cones. Probably it is to this action that we ma}' attrib- 



