844 BULLETIN UNITED STATES GEOLOGICAL SUEVEY. 



sufficient time is afforded for the completion of the chemical change. 

 Such material as was most readily transportable during the period of 

 formation of the conglomerate is, by reason of its lighter specific gravity, 

 comparatively free from the coloring ingredient. In one feature, per- 

 haps, may this conglomerate be regarded as exceptional if compared 

 with others. This consists in the irregular accumulations of physically 

 differing masses. Irregularly shaped masses of finegrained, loosely 

 cemented material may be regarded as inclusions within the normal con- 

 glomerate. Their existence is taken advantage of by erosive forces, and 

 they rapidly yield to the oft-repeated attacks. 



Within the various groups exhibited on South Eiver, the process of 

 their formation could be most admirably studied. Erosion by flowing 

 water, assisted probably by the movements of temporary glaciers, have 

 first shaped the general outlines of the valley. Thus was the conglom- 

 eritic deposit cut apart after a portion of the hard trachyte protecting 

 it had been removed. Subsequent flows, more particuarly from the 

 high ground east of the valley, cut parallel gorges and ravines into the 

 readily yielding conglomerate. These had a trend at approximately 

 right angles to the course of South Eiver. The ridges, formerly 

 dividing them, now remain, in consequence of still further denudation, 

 as the transverse walls above mentioned. Their relative position to the 

 main cliffs eastward supports this view. While most likely fluviatile 

 erosion determined the first great separations of the mass into groups, 

 other agents were employed to carve out the individual forms. From 

 observations made on the ground, it would appear that the walls were 

 slowly growing thinner, owing to the gradual separation of columns 

 from their sides. Among the most potent agents preparing absolute 

 removal of material, we must count the influence of frost. During prob- 

 ably eight months of the year the temperature falls below the freezing- 

 point at night, while during more than one-third of the time the heat of 

 the day will produce a complete remelting of the frozen water. Wher- 

 ever, then, we have loosely cemented material, readily saturated by 

 water, we will find that the repeated expansion upon freezing eventually 

 places the component particles of rock in such positions as to be easily 

 removed. It was observed that innumerable bowlders of varying sizes 

 projected from the sieep walls. Precipitated moisture, finding its way 

 down along the steep surface, will encounter one of these obstacles, and, 

 concentrating its volume along one line, will follow down along either 

 one side or the other of the erratic block. Thus gradually a groove will 

 be eroded downward from either side of the bowlder. If we continue 

 this process for a long period of time, it must finally result in an isola- 

 tion of a columnar mass, with the bowlder as a protecting cap. Exami- 

 nations showed that this method of formation would satisfactorily 

 explain not only the form, but more particularly the distribution, of the 

 monuments. They occur most densely clustered along the base of the 

 walls and along their edges. Again they closely stud the sides of newly 



