ON SEISMOLOGICAL INVESTIGATIONS. 



99 



If we omit the first two entries in this table it appears that the 

 average number of earthquakes in any period is approximately directly 

 proportional to angular deflections of the pole-path during that period. 



A similar result is shown in the three group averages which are 

 given : — 



Group II. 



In this instance, where the path of the pole relatively to its mean 

 position has been fairly uniform, earthquake-frequency does not appear 

 to have been influenced. 



"VII. On the Change of Level on two sides of a Valley. 



It has been found that under certain but frequently recurring con- 

 ditions the two opposite sides of a valley move in opposite directions 

 at the same time. On bright fine days the inclinations of the sides 

 of a valley decrease. At night they increase. A valley may therefore be 

 supposed to open and close. These conclusions, which do not necessarily 

 apply to all valleys, are based on observations taken in two very difie- 

 rent localities. The first were made in Tokyo, Japan, by means of hori- 

 zontal pendulums giving continuous photographic records, installed on 

 the two sides of a valley cut in alluvium.^ 



The second series were made on the two sides of a valley cut in chalk 

 at Shide, near Newport, in the Isle of Wight. On the western side of 

 the valley the instrument employed was an astronomical level reading 

 to l"-0 of arc. On the eastern side change of level was continuously 

 recorded by a horizontal pendulum easily reading to 0"-5 of arc. Both 

 were well founded and well protected from direct effects of solar radia- 

 tion. When the instruments were side by side they gave similar and 

 practically identical results. 



The level, which was in charge of Mr. H. G. Morgan Hobbs, resident 

 at Sunnyside, was read three times a day from July 26 to August 27 

 1905. The difference between successive readings usually varied between 

 0"-5 and 3"-0, and with but few exceptions these readings indicated 

 changes in level in opposite directions. During wet weather the diurnal 

 movement was eclipsed by a rapid movement in a direction which corre- 

 sponded with a closing of the valley. This sometimes amounted to 18"-0. 



These diurnal changes in the vertical have been found in chambers 

 excavated in rock, and at other installations when the daily change in 

 temperature has not exceeded 2" F. They have been recorded in the^New 

 Red Sandstone at a depth of 19 feet. From these and other observations 

 it is clear that they cannot be attributed to any warping effect in the 

 instrument or change in temperature at the base of a pier on which an 



See Brit. Assoc. Rep., 1896, p. 132. 



H2 



