Earthquake waves through the earth. 359 



coexist in a rigid mass that the phenomena have been held to 

 prove a high rigidity of the interior 1 . If however the present 

 hypothesis be valid that argument loses its force. 



Let us now enquire what conclusions may be drawn from the 

 present hypothesis regarding the constitution of the magma. 



The velocity of the wave of expansion of the liquid alone will 



be a/ j\, and if this is the wave which produces the preliminary 



tremors then \/ j] = H km. per second. 



/p 

 The velocity of the gaseous wave will be a / —f. . If this is 



the wave of the second observed type its velocity is about 5*29 km. 



P 



per second. This will give — ~ = (5*29 km. per second) 2 . 



Reducing this to feet, since 1 km. = 0*621 mile, 

 -^ = (5-29 x 0-621 x 5280) 2 . 



Here P is the pressure on the magma in its saturated con- 

 dition. We may take it as fairly certain that the magma is 

 normally saturated at the bottom of the crust, otherwise a layer 

 of gas would be formed in that position, which, though perhaps 

 locally occurring before explosive outbursts like that of Ban- 

 daisan 2 , is as a rule highly improbable. Hence P may be 

 considered to be the pressure of the crust and D the density of 

 the magma under it in a saturated condition. 



Let then w be the mass of a cubic foot of water, p the specific 

 gravity of the crust, k its thickness in feet, and g the accele- 

 ration of gravity. Then the pressure arising from the weight of 

 the crust will be pkivg = P. 



Assuming the thickness of the crust to be 25 miles, and its 

 specific gravity 2"68, we shall then have, 



2-68 x 25x^280 x 32 x w _ ^ x ^ ^ ^ 



which gives rD = 0-037627w. 



Assuming further that the density of the magma is three times 

 that of water, which is the case with a heavy basic rock, we have 



r 3tu = 0-037627w, 



and r = 0-01254, 



1 Milne's Seismology, pp. 115, 116, Also Darwin's Tides, p. 237, 

 - Nature, 13 Sept. 1888. 



