part 2] IN THE FREQUENCE OF EARTHQUAKES. 105 



quarter of that of the solar semidiurnal tide, but this has a node in latitudes 



, 1 

 + sin /- — not far from Italy and Japan. From the fact that the 



difference between day and night persists throughout the year without 

 change of sign, only of amplitude, I should conjecture that the explanation 

 is partly thermal and partly tidal, although I admit that I do not know how 

 heating could produce any effect at such depths, unless it could penetrate in 

 some way down the fault-planes. It is very difficult, for instance, to exclude 

 the direct effect of the surface in finding temperatures in borings. 



If the diurnal bodily tide is important in Japan, then the solar semidiurnal 

 tide should produce a far greater effect at the Equator and in high latitudes, 

 and the lunar semidiurnal tide a greater one still ; the action of the two 

 together would give a strong fortnightly term. The absence of these would 

 indicate that the effect is mostly thermal. 



Mr. W. H. Booth suggested that the Earth's temperature, 

 which increases about 1° Fahr. per 50 feet of depth, was due solely 

 to the friction set up by the work of the sun and moon in bending 

 the Earth's outer surface, and that the temperature probably attains 

 a maximum at a comparatively small depth. This is supported by 

 the high temperatures found in boring the Simplon Tunnel, when 

 boiling water was met with several thousand feet above sea-level. 

 Mountains, of course, standing above the general level, are particu- 

 larly subject to tidal action. All work is dissipated as heat, and, 

 when the rock-temperatures attain a certain limit, the rocks 

 may be so weakened as to give way : as the Author of the paper 

 suggests, the tendency would be to give way at times of maximum 

 effort of tide-producing agents. 



The Author, replying, said that he was by no means satisfied 

 that the variations in frequency were really consequences of the 

 hypothesis on which they had been discussed. All that could be 

 said was that the results were quite concordant and in agreement 

 with the supposition that the tidal stresses set up by the sun were 

 not without effect on the frequency of earthquakes. Some of the 

 results were of interest and value apart from the truth or otherwise 

 of .the hypothesis, which was being tested by a discussion of the 

 shocks by lunar as well as solar time. 



