648 J. V. LEWIS — OHIGIN OIF PILLOW LAVAS 



another, they tend to assume somewhat rounded or domed upper surfaces 

 and molded bases. Hence the pillows at the bottom of a flow are apt ,to 

 be flat beneath, unless a soft silt or ooze permitted some degree of sink- 

 ing into the sediment and the development' of curved lower as well as 

 upper surfaces. Those higher in the bed tend to fit more or less per- 

 fectly on the underlying ones, the degree of adaptation varying inversely 

 with the promptness with which a restraining integument is developed. 

 It is particularly to be noted that according to this view the pillows im- 

 mediately assume fixed positions as they are formed. There is no rolling 

 over and over on one another, as so many have supposed, nor intersphe- 

 roidal motion of any kind involved in their production; neither are the 

 pillows pressed together while still plastic, since only those in process of 

 formation at the surface are soft and yielding; 



From the mode of construction it necessarily follows that the various 

 axes of the masses under the influence of gravitation assume positions 

 approximately parallel to each other, the longest marking the direction 

 of flow and the shortest the vertical at the time of their formation (see 

 plate 19, figure 1, and plate 23, figure 1). Other conditions being equal, 

 the steeper the slope the more elongated and bolster-like will be the pil- 

 lows. An increased pressure and excessive flow of lava would enlarge 

 the bulbs so rapidly that they would flatten out and tend to merge into 

 the sheet-like pulsations which Button has described as characteristic of 

 pahoehoe. A similar result would follow on a considerable increase in 

 temperature, since this would delay the formation of the tough capsule 

 on which the rounded shapes depend. As pointed out by Capps, the 

 upper surface of a bed of pillows consists of ^^a succession of domes re- 

 sembling the surface of a magnified cobblestone pavement." ^*^ 



Several observers have noted the general parallelism of the longest 

 axes of the pillows, and the further fact that they are approximately 

 parallel to the bedding-planes of accompanying sedimentary rocks, but 

 no effort has been made to interpret these relations. Under any suppo- 

 sition of rolling, whether in a surface flow or by the action of intense 

 dynamic forces beneath the surface, the longest diameters would neces- 

 sarily mark the direction at right angles to that of the movement." Capps 

 has emphasized the value of the sharp distinction between the top and 

 bottom surfaces of a sheet of pillow lava in the interpretation of a com- 

 plicated structure, but it is possible for these distinguishing characters 

 to be greatly modified, and they may even disappear altogether. For 



1*0 S. R. Capps : Manuscript report on the Ellamar district, Alaska, U. S. Geol. Survey, 

 to which I am permitted to refer in advance of publication. 



