ON THE EARTHQUAKE AXD VOLCANIC THENOMENA OF JAPAN, 131 



An inspection of this table shows that, although the pendulums A and 

 J were separated by a distance of about 270 yards, and were differently 

 installed, there has been a general agreement in the direction of their 

 displacements, and that this is particularly marked Avhen the movements 

 of A were decided. When there has been disagreement in the direction 

 of motion, the movements of A have usually been small, and it is not 

 unlikely tliat such discrepancies would disappear if a time had been taken 

 to represent the daily mean position farther removed from the hour at 

 which the western movement of the diurnal wave commences. The 

 changes indicated in the table are also shown diagrammatically in con- 

 junction with a curve showing the fluctuations of the barometer (p. 129). 



What is true for A and J is generally true for J and K. Although 

 between October 13 and November 20 there are six decided barometrical 

 fluctuations, and during tlie same interval there are six decided westerly 

 movements of J, the crests and depressions of these diagrams do not 

 retain the same relative position. For example, on October 1 7 the 

 barometrical crest corresponds to the westerly extension of J, while on 

 November 5 a similar movement of J corresponds to a barometrical 

 depression. Although at times it appears as if there were a close relation- 

 ship between barometrical fluctuation and the movements of the pen- 

 dulums, the diagram (fig. 4) indicates that, although both phenomena are 

 nearly identical in having a periodicity of between two and seven days, it 

 does not show that they are absolutely synchronous. It seems that the 

 instruments at J and K, which were within six feet of exposed ground, 

 M'ith or after rain moved westwards, but equally large westerly motions 

 have occurred without i-ain. 



g. On the Diurnal Wave. 



In considering the results towards which the observations of tlie 

 daily wave point, it is necessary to consider the observations made during 

 the past year in conjunction with those described in the Beport for 

 1893-94. 



When instruments were installed upon the rock in caves as at Kama- 

 kura (C and D), at Kanagawa (I), and in Yokohama (G and H) the daily 

 •wave was not perceptible. This by no means precludes the possibility of 



K 2 



