B. A. Daly — Mechanics of Igneous Intrusion. 279 



His fusion curves show that, for the same increase of pressure, 

 liquid naphthalene gains in specific gravity about twice as fast 

 as solid naphthalene. The compressibility of a fused silicate 

 rock is probably, then, about twice that of the same rock when 

 solid. But his diabase fusion curve demonstrates that the ther- 

 mal expansibility of the liquid rock is about 1-9 times as great 

 as that of the solid rock. Thus a block of cold, solid gabbro 

 immersed in a deep-seated molten magma of the same chemical 

 composition, would be less condensed by the pressure than the 

 molten rock, but the effect on relative densities would be 

 partly compensated by any superheating -of the magma. 

 It is, moreover, clear from the facts of cols. 1, 3 and 4, 

 Table IY, that, at 1400° C, this different thermal expansi- 

 bility would occasion but a small difference in the respective 

 ratios of the specific gravities of a solid rock at ordinary tem- 

 perature and at 1400° C. to the specific gravity of a magma at 

 the latter temperature. The same is doubtless true of com- 

 pressibility, since that property- bears a simple relation to 

 expansibility. The compressibility of glass and of crystalline 

 silicates is known to be very low. The compression suffered 

 by glass, for example, is about '0000026 of its volume for 1 atm. 

 The weight of even 10,000 meters of rock with an average 

 density of 2*75 would cause a density increase of much less 

 than one per cent in glass. It is therefore probable that the 

 difference of density between magma and immersed block 

 would not be affected through pressure, at the great depth of 

 10 kilometers, by as much as 1 per cent of the density of either 

 one. 



Relative densities of solid and molten rocks under plutonic 

 conditions. — It is perhaps necessary to state that any errors 

 introduced by the application of the expansion coefficients of 

 Barus and Reade to rocks differing from their types, cannot 

 seriously vitiate cols. 3 and 4 in the table ; and that the relative 

 differences of density signalized in cols. 2 and 3, or 2 and 4, 

 Table IY, will persist practically unchanged, whether the solid 

 rock is immersed at a temperature near its melting point or a 

 couple of hundred degrees above it. 



Not with absolute certainty, then, but with high and cumu- 

 lative probability, we are entitled to use cols. 2 and 3, or 2 and 

 4, of Table IY as expressing the essential contrast of densities 

 between plutonic magmas and fragments of solid rock rifted 

 off by them from the walls or roof of their respective chambers. 

 The following conclusions are thus possible : 



1. A solid block of each igneous rock would sink in its own 

 molten glass. 



2. Blocks of the basic eruptive rocks would sink in all of 

 the magmas except in a very basic peridotite. 



