646 J. V. LEWIS ORIGIN OF PILLOW LAVAS 



A Theory of bulbous Budding 



GENERAL DISCUSSION 



A comparison of the descriptions and illustrations of pillow lavas from 

 various parts of the world with the excellent examples in the Watchung 

 Mountain basalts in Xew Jersey shows that in all essential characters 

 they are identical. TTith the advantage of the numerous studies and 

 discussions of origin that have been published, and especially the illumi- 

 nating direct observations of Green, Anderson, and Day, it should now 

 be possible to devise a satisfactory theory to account for the production 

 of these forms without resort to strained or improbable assumptions. 

 This I have attempted to do in the following paragraphs. 



1. HIGHLY LIQUID LAVA 



Since pillow structure is found only in free-flowing basalts and closely 

 related lavas, it is evident that any high degree of viscosity must serve 

 effectually to prevent its formation. Hence it becomes necessary to 

 reject such characterizations as ^Tiighly viscous pahoehoe," which some 

 authors (including myself) have used in former years to describe this 

 structure. On the contrary, the necessary physical condition of the lava 

 is rather its capacity for retaining a high degree of liquidity through a 

 relatively long period of cooling and the development of a notable degree 

 of viscosity only mthin a limited range of temperature as it approaches 

 rigidity. 



2. SMALL CONTINUOUS SUPPLY 



In the declining stages of a large flow, when the lava lies nearly mo- 

 tionless, the formation of small cracks and orifices in the cooling crust 

 relieves the internal pressure, and hence checks still further the forward 

 motion by permitting the escape of liquid lava at the front and along 

 the sides of the flow or, in case of sufficient slope or internal pressure 

 from any cause, even on its upper surface. Thus the process may sub- 

 stitute a multitude of small flows for the one great flow that has become 

 retarded or stopped in its course, or in the case of slow gentle extravasa- 

 tion of lava from the beginning the multiple-flow principle may come 

 into play at once. In either case the continuation of the process depends 

 on the maintenance of favorable conditions. Thus it may be terminated, 

 for example, by the failure of the supply of lava or the increase of the 

 supply to the point where large flows would be inaugurated (or reestab- 



