Professor J. Milne — Earthquakes of Japan. 85 



there is a light metal plate with a slit in it parallel to the length of 

 the boom. Underneath this floating slit, but at right angles to it, 

 there is a narrow slit in the top of a box. Light passing through 

 the two slits goes into the box as a point, which is received on a 

 drum carrying a photographic film. These instruments, which are 

 usually arranged in pairs, are placed to point N.W. and N.E. or 

 parallel and at right angles to the strike of a distant range of hill. 

 Two have been used at Kamakura on the solid rock, and two at 

 Yokohama and one at Kanagawa on the soft tuff rock. These instru- 

 ments were in caves. Also I have had two underground on the 

 alluvium and one at my house on a solid stone column in Tokio. 

 The movements recorded have been as follows : — 



1. The Wandering of Pendtdums. — All the horizontal pendulums, 

 wherever situated, have slowly wandered from their normal position. 

 Those situated on the rock have often gi'adually moved to one side 

 and then returned, the double excursion usually taking from two 

 days to a week. These wanderings might be due to a local warping 

 of the supporting column; but inasmuch as it has generally happened 

 that the periods of great movement and of comparative rest have 

 coincided in time, it would seem that the movements ai-e in all 

 probability due to a more general cause. Because certain move- 

 ments have usually been marked (but by no means always) at or 

 after a rainfall, it seems possible that they may be connected with 

 fluctuations in the volume and flow of underground water, — the 

 pendulums nearest to this water moving the most. 



2. Daily Waves. — In no instance have I observed a diurnal move- 

 ment of the pendulums at stations situated on the rock. In Tokio 

 the movements occur underground and on the surface ; they happen 

 at the same time and they are proportional in magnitude. At my 

 house, for example, at about 10 a.m. a tilting commences on the 

 N.E. side, and it reaches a maximum of one or two seconds of arc 

 about midnight, when a sinking takes place until about 6 a.m., when 

 the pendulums remain fairly steady for some hours, after which they 

 again commence to rise. Because it was observed that the move- 

 ments underground were greater than those recorded on the surface, 

 it seemed possible that they might find an explanation in the fact 

 that the underground instruments were nearer to pervious strata in 

 which water fluctuated than the instruments installed in my house. 

 By quickly emptying a well in my garden of about two tons of 

 water which was run off down a hill, the pendulum on a column 

 (10-i feet distant) behaved as if the ground had been relieved of 

 a load and therefore had risen on the well side. Because the 

 direction of movement of the pendulums underground and on the 

 surface is away from the side from which during the day the greatest 

 load is being removed by evaporation, it is not unlikely that the 

 main features of the diurnal wave may be due to this cause. 



3. Earth Pidsations (Tremors). — In Italy I understand that earth 

 tremors are as marked underground on the rock as they are on the 

 surface. In Japan, on three underground rock foundations, I have 

 not observed a single case of earth tremors. Both underground and 



