104 MR. R. D. OLDHAM ON A SEASONAL VARIATION [vol. lxxiv, 



the negative sign of the fourth term being used for the summer, 

 and the positive for the winter, half of the year. 



The Japanese record, treated similarly, gives the formula 



P=l-0.i--10sin. (t + 107°15') + *04 8in. (2i + 130°), . . . (6) 



which, compared with (3), shows amplitudes of about one-third 

 of those in the Italian record, and epochs which precede midnight 

 by lh. 9m. and Ih. 20m. respectively. The meaning and interpreta- 

 tion of this difference being at present under investigation, further 

 reference to it must be deferred.] 



EXPLANATION OF PLATE XII. 



Diagram illustrating the diurnal and seasonal frequency of earthquakes 



in Italy and Japan. 



Discussion. 



Dr. J. W. Evans congratulated the Author on the clearness of 

 his exposition of the intricate relations with which his communica- 

 tion dealt. He seemed to have established satisfactorily the reality 

 of the diurnal and seasonal variations in earthquake frequency for 

 which he contended, and their relation to the diurnal inequality of 

 the solar tides in the earth's crust. One naturally compared these 

 results with the pendulum observations by Hecker at Potsdam in 

 a chamber excavated at more than 80 feet below the surface of the 

 ground. These indicated variations in the conformation of the 

 earth's crust, which, like those in the frequency of earthquakes, 

 must be due to stresses of a periodic character. The diurnal 

 variation is, however, attributed to changes of temperature, partly 

 because it is too large in comparison with the semidiurnal and 

 lunar variations to represent the diurnal tidal inequality, and partly 

 because it is greatest in summer and least in winter. It is difficult 

 to understand how the variations in temperature which mainly occur 

 in the outer 6 or 8 inches of the earth's crust and have practically 

 ceased to exist at a depth of 2 feet could have such a result ; but 

 there is no other known cause that can be suggested, except the 

 light pressure of the sun's radiation, and that only amounts to 

 75,000 tons for the whole hemisphere exposed to sunlight. The 

 diurnal variation in earthquake frequency described by the Author 

 cannot be due to the daily changes of temperature, for it is least 

 in summer when these are greatest. 



By permission of the President, Dr. Evans then read the follow- 

 ing extracts from a letter by Mr. Harold Jeffreys, who had 

 seen the abstract, but was prevented from being present at the 

 discussion : — 



The dynamical explanation of these observations is likely to be somewhat 

 difficult. The diurnal effect (period 24 hours) might arise, either from the 

 diurnal bodily tide, or from heating. The former is rather small, depending 

 on the inclination of the Equator to the ecliptic, and should have opposite 

 signs in winter and summer at the same plane, and in the Northern and 

 Southern Hemispheres at the same time. Its amplitude is only about a 



