﻿Vol. 64.] OKIGIN OF THE PILLOW-LAVA IN CORNWALL. 267 



sections exposed upon joint- faces, and each is seen to consist of a 

 series of alternating concentric bands of more or less amygdaloidal 

 rock. The amygdules are smallest in the outermost layers, and 

 decrease in number but increase in size towards the interior. The 

 centre is often highly vesicular, with a large cavity from which the 

 infilling ealcite has been weathered out. Some of these central 

 cavities occui^y a large area of the section, measuring as much as 

 2 feet in diameter in a pillow 6 feet in diameter. In short, each 

 pillow is a gigantic thick-walled bubble, blown out by the included 

 steam and other gases, and showing the stretching of the walls in 

 the concentric bands of drawn-out vesicles which occupy the outer 

 part of its substance. 



Formation o£ the Pillows. 



Though the inflation of the pillows seems so obvious in the field, 

 we could not for a long time understand how it happened that this 

 inflation took such a peculiar form. Why did not the whole sheet 

 expand into a thick mass of vesicular pumice, instead of forming 

 these characteristic spheroids ? 



While studying the lava, it struck us that the pillows were so 

 spongy and full of cavities, that before the vesicles were infilled 

 with ealcite the pillows must have been very light. We therefore 

 attempted to measure the cubic contents of rock and interspace in 

 the larger pillows. This proved to be more difficult than we had 

 anticipated. Photographs show numerous sections through sphe- 

 roids ; but unless the section happens to be taken exactly through 

 the central cavity, we must underestimate the total bulk of the 

 cavities. Tangential sections miss the central cavity, and may only 

 pass through the outer and closer zones. Perhaps no one section 

 that we have measured passes exactly through the centre of the 

 mass, but we have selected two, one of which is figured (PI. XXYIII), 

 as representing fairly well the usual structure. 



The obvious and simplest method of measuring the relation of 

 vesicle to solid rock would seem at first sight to be by decalcification 

 of a whole pillow, so as to remove the infilling of the vesicles, and 

 allow a direct comparison of the weight before and after the ealcite 

 was removed. Eut apart from the practical difficulty of dealing 

 with a mass containing 10 or more cubic feet, there was the obvious 

 objection that some part of the carbonate must be a decomposition- 

 product of the igneous rock, and to remove the whole would lead 

 to underestimation of the original weight. Also, some of the 

 vesicles are filled with silica, others with chlorite. We have, 

 however, measured the approximate area of vesicle and rock in 

 successive shells in the pillows, and have calculated the cubic 

 contents of each in the different shells seen in the sections. 



We do not pretend that the result is more than a rough approxi- 

 mation ; but it is very curious. Taking the original specific gravity 

 of a lava of this sort as about 2*7, cavities occupy nearly one-half 

 of the mass, reducing the specific gravity of the whole pillow, in the 

 two examples measured, respectively to 1*40 and 1*45. 



