76 H. C. COOKE 



Pillow structure. — Pillow structure is well developed in the lavas 

 over a large portion of the region, more particularly in the andesitic 

 types, which commonly have large pillows from two to three feet 

 in length. The basaltic lavas for some reason do not seem to 

 form large pillows, the maximum observed being one foot in length; 

 and on account of the blackness of the basalt it is more difficult to 

 recognize and examine the pillows except on unusually clean, wave- 

 washed surfaces. 



A section made from top to bottom across an andesite flow 

 about 130 feet thick on Windy Lake showed the following arrange- 

 ment: At the top there is the usual narrow, schistose zone, below 

 which there is a zone of pillows about 70 feet in width. At the 

 top the pillows are from two to three feet in length ; they gradually 

 decrease in size, till near the middle of the flow they attain their 

 minimum length of ten inches to one foot. Here the pillow struc- 

 ture ceases rather abruptly and passes into a fragmental zone 

 about six feet in width. This is a true volcanic breccia, made up of 

 fragments of fine-grained lava up to six inches in diameter imbedded 

 in a matrix of lava of somewhat coarse grain. Some of these 

 fragments are encircled by a whorl of lava about an inch in width, 

 in which flow textures are prominent, as if the fragment had been 

 revolving in the viscous lava. Below the zone of breccia, fine- 

 grained massive lava occurs. The grain of this gradually increases 

 in size as far as the bottom of the flow, where a grain of approxi- 

 mately I mm. is attained. This massive, relatively coarse-grained 

 lava was observed in contact with the glassy, ellipsoidal surface of 

 the adjacent underlying flow. 



Only one perfect section across the whole width of a thick flow 

 was observed, but a large number of partial sections were obtained, 

 and the data are sufficient to show that the above represents the 

 section wherever the flow is sufficiently thick to possess a massive 

 base as well as a pillowed top. The succession described may 

 therefore be used to indicate the position of the top of a tilted flow, 

 even where only a part of the flow is visible. Of course this method 

 is useless in flows pillowed throughout their entire thickness. 



The method of determining the position of the upper surface 

 of the flows from the flattening of the pillows, described by Daly, 



