652 J. V. LEWIS ORIGIN OF PILLOW LAVAS 



prime importance. The mobility hinges on both the composition and the 

 temperature and will be greatly affected by the kinds and proportions of 

 both the mineral-forming and the volatile constituents. 



11. CONTACT WITH WATER 



The influence of water on a subaqueous flow or one entering a lake or 

 sea can not yet be specified with any considerable degree of confidence. 

 Possibly characteristic differences between subaqueous and subaerial flows 

 exist, but the principal features developed under the two conditions are 

 indistinguishable in the present state of knowledge. Presumably the 

 rate of cooling is somewhat accelerated by immediate contact with water, 

 at least in certain stages of the process, as at the initial budding of each 

 new bulb or pillow, and this probably has a modifying influence on the 

 size, shape, and other characters of the completed individuals ; but to a 

 great extent these must depend also on the initial temperature of the 

 lava of each diminutive outbreak, the pressure under which it flows, its 

 chemical constitution, and other indeterminate factors. 



Contact with water may also have the effect of either hindering or 

 promoting the formation of the pillow structure itself. Conditions of 

 temperature and flow that are favorable to the production of extensive 

 pillow lava on land, for example, may become so modified by flow- into 

 water that the extent of the structure will be much restricted. On the 

 other hand, it is conceivable that a flow of such volume and fluidity as 

 would spread rapidly into a continuous sheet on land might be checked 

 and largely transformed into pillow lava on entering a body of water. 

 Hence it seems safe to conclude that neither th'e presence nor the absence 

 of water, per se, can be predicated as particularly favorable to the forma- 

 tion of this structure. 



12. INTRUSIVE PILLOW LAVA 



Direct intrusion of lava into soft oozes or loose sediments on the sea- 

 bottom in general would differ from subaqueous eruption only in the 

 slight resistance to flow offered by the sediments and in less rapid cooling, 

 perhaps, owing to the absence of convection currents. In other respects 

 such an injection would be essentially an extrusive flow. Indeed, a sub- 

 aqueous eruption on a sea or lake bottom of such insubstantial character 

 would of necessity penetrate the underlying sediment in part, and like- 

 wise a lava stream from the land would sink into a semiliquid ooze or 

 quicksand on entering a body of water where such conditions prevail.. 

 While, strictly speaking, either of the foregoing types must be called in- 

 trusive, the controlling condition is manifestly the same in both, namely, 



