110 KEPORT — 1895. 



in Tokio or Yokohama, the records from this place are extremely few. 

 Those which have been obtained, extending over two or three days, show 

 straight lines like fig. 1, Plate II. There are neither daily curves nor tremors. 

 From Kanagawa, although the cave is very wet and the conditions for 

 observing very unfavourable, for about two months everything has worked 

 satisfactorily. Like the Kamakura records they do not show tremors or 

 daily waves, but they do show unfelt earthquakes and wandering. For 

 example, on May 5 the boom moved as if by a N.E. tilting as much as 

 14". This it reached on May 7. From this date it slowly returned to its 

 starting point, which it reached on May 12. Small shocks occurred on 

 the 2nd, 4th, and 6th. 



(q) Conclusions. 



Inasmuch as the analysis of materials already accumulated is not yet 

 completed, and as certain experiments require to be repeated or amplified, 

 it is premature to formulate definite conclusions. All that can therefore 

 be done is to outline the form which conclusions may possibly assume. 



Although I understand that Italian observers have found that tremors 

 are as marked underground, even on the rock, as they are on the surface 

 in Japan, this seems to be only true for the alluvium. Underground on 

 the rock at three stations, with such instruments as I have employed, 

 there has not been even an indication of tremors. Neither have 

 daily waves been observed. All the pendulums, whether on the rock 

 or on the alluvium, from time to time leave their normal position, 

 moving for two or three days in one direction and then slowly returning. 

 These movements, which have been called wanderings, sometimes indicate 

 a tilting of as much as 14''. Because these movements have often been 

 accompanied by local earthquakes, it seems possible that they may 

 actually represent rock bending, the earthquakes announcing the fact 

 that resistances to the process are being overcome. 



Some of the wanderings noted on the alluvium may possibly be 

 attributed to disturbances in the subterranean circulation of water after 



rainfall. 



Although the daily movement of the pendulums has been most marked 

 by those which are nearest to water level, because they only show a single 

 wave during the day, while the water in a neighbouring well rises and 

 falls twice during the 24 hours, the daily wave cannot altogether be 

 attributed to the movement of subterranean water. Because certain 

 diao-rams have shown a superimposed wave, it is possible that the cha- 

 racter of the daily wave may now and then be influenced by subterranean 

 water. Because a wave may be produced by relieving an area in the 

 vicinity of a pendulum of a load, as, for example, by taking 2 tons of water 

 out of a well which is 104 feet distant from pendulum A, and pouring the 

 water away down a slope, it seems likely that the daily wave is produced 

 by an action of this description. The action suggested is that which 

 takes place every day when the sun shines or the wind blows across 

 <n-ound which is open and that which is covered, for example, by forests or 

 buildino-s. By evaporation one area is rapidly relieved of a load, while 

 the adjacent area loses but little. For example, experiment shows that 

 on fine days an open grass-covered area 140 feet square in front of my 

 house, which is 120 feet long and runs E. and W., loses in 12 hours about 

 5 tons of moisture. At the back of the house, where the ground is 

 sheltered from the sun, evaporation is small. As confirming this view it 



& 



