PYEOOLASTIGS OF HEMLOCK FORMATIOi^. 141 



intevbedding' of layers in which coarse and finer fragments prevail, illus- 

 trating well the varying intensity of the volcanic discharges. 



Owing to the fragraental nature of the exposures, it is impossible to get 

 a correct idea of the maximum thickness of any of the tuff deposits. 

 Exposures were seen in the north half of sec. 5, T. 43 N., R. 32 W., which 

 gave a thickness of over 500 feet for some of these deposits, but as their 

 farther continuation had been cut off by valleys, most probably eroded in 

 the tuffs, no means was afforded of determining their total thickness. 



It is almost needless to state that the most of the tuffs have undergone 

 a great amount of alteration. The alterations were apparently due to an 

 interchange of the various elements without any essential variation in the 

 chemical nature of the rock as a whole. Since water is the chief agent 

 through which alterations occur, these always begin along the interstices. 

 In the case of the fragments the alteration accordingly proceeds from the 

 outside inward, and ordinarily at an equal rate all around the fragments, 

 following its contours. In this way zones of somewhat different mineralogical 

 composition are formed, surrounding the less altered part of the fragment. 

 This secondary zonal structure may be obsei'ved more or less imperfectly 

 in almost all of the sections made from the breccias, but is much better 

 shown in the field, where the concentric zones are well brought out on the 

 large weathered surfaces of the bowlders. 



In each case the outside, lighter-colored zone is chiefly made up of 

 chlorite, from which project light-green hornblende needles into the matrix 

 beyond. Less commonly we find it composed of epidote grains and chlo- 

 rite. Inside of this zone the mineral elements composing the fragments 

 sometimes can not be determined with any great degree of certainty. 

 Where determinable, the alteration products are found to be the same as are 

 produced from the corresponding rocks in the lava flows. As also in the 

 lava flows, some of the fragments of the denser rocks have become almost 

 opaque from the quantity of minute secondary epidote and sphene grains. 

 These have a lighter green color than the less altered fragments. 



In examining many of the tuffs, one is repeatedly struck by the large 

 amovmt of space occupied by the cementing material. In some cases cavi- 

 ties of very considerable size were left between the fragments. It appears 

 that the fragments must have been lying very loosely. This fact tends to 

 confirm the eolian origin of the rocks, since water deposition tends to bring 



