— 
ON SEISMOLOGICAL INVESTIGATIONS. 101 
sudden contraction of from 3 to 4 per cent.° Beneath an ocean bed 
with a gradient of 1° I’. in 20 feet we should expect this to take place 
at a depth of about eight miles, but beneath a continent with a gradient 
of 1° in 60 feet at a depth of about twenty-four miles. 
XIII. A Possible Cause of Meyaseismic Activity. 
That earth rest after megaseismic activity is roughly proportional 
to that activity as measured by the number of large earthquakes in a 
group (p. 24), and that activity in the world is most frequently 
repeated after fifteen or thirty days of rest (p. 24), suggests that the 
cause which brings large earthquakes into being which cannot be 
traced to epigenic influences may be due to the steady dissipation of 
earth heat. In the first place, this view finds strong support in the 
fact that the regions where geothermic gradients are steepest are those 
from which megaseisms most frequently radiate. 
Voleanic rock, when passing from the fluid state to the solid, con- 
tracts suddenly (see p. 32), and something similar happens when 
molten slag solidifies. Information bearing on this subject was very 
kindly obtained for me by Mr. J. J. Shaw, of West Bromwich. To 
get rid of the slag from an iron furnace it is run into moulds or holders. 
As it mounts upwards in one of these, its outer edges are seen to con- 
tract or curve inwards, leaving a small space between the side of the 
holder and the hot ‘metal.’ The hot stream, as it continues to pour, 
fills up this space. When, however, it has reached a height of one 
and a half or two inches more in the holder, a second contraction 
occurs. This intermittent contraction and filling up the space it has 
left goes on until the holder is full. ‘ When the block is turned out 
it shows striz round its sides which correspond to the intermittent 
solidifications. Although the conditions of a cooling block of slag are 
different from those of a cooling globe, they suggest a series of spasmodic 
contractions at regular intervals rather than a contraction that is 
uniform,’ a phenomenon which is roughly illustrated in the successive 
sequence of large earthquakes. 
When the block cools it frequently cracks, and hot material is 
exuded. This is due, as pointed out by Mallet, to the grip of the 
contracting outside shell upon the hot interior. 
The huge dykes filled with volcanic rock which traverse many 
countries, together with the fissure eruptions which have buried many 
thousands of square miles to depths of from 2,000 to 6,000 feet of lava, 
correspond, but on a gigantic scale, to the phenomena observed on the 
surface of the cooling slag. 
With each sudden yielding vibrations or waves would be generated 
on the surface of the viscous mass, and if it is assumed that this is 
homogeneous, these would be propagated beneath the crust at a uniform 
velocity, which is the case with the large waves of earthquakes. 
The suggestion here made is the reverse of the old idea. It is not 
a nucleus that contracts to leave a shell to follow downwards, but a 
* See Bulletin of the U.S. Geolog. Survey, No. 103, ‘Igneous Fusion and 
Ebullition,’ by Car] Barus. 
