ox THE EARTHQUAKE AND VOLCANIC PHENOMENA OF JAPAN. 133 



especiiilly upon frosty niglits, were extremely erratic, and they may 

 probably have been produced by the freezing of the ground. When the 

 cold was not great the movements were small, and here, again, the ground 

 around the instrument was so well shaded by trees that after a snow- 

 storm it would take from ten to fifteen days to melt away the snow, which 

 at other places disappeared in one or two days. At Q, R, S, and O 

 the movement during the day was towards the side on which the ground 

 near to the instruments was most exposed to the sun. 



One very important observation especially marked at stations where 

 large diurnal waves were recorded was that on rainy or doiidy daya these 

 instruments ivere steady, and no dbirnal wave ivas recorded. 



The general conclusion to which these observations on the diurnal wave 

 point is that on the alluvium on open ground, the daily wave is most 

 pronounced on the surface, it is less in amplitude, but it may be decided at 

 a depth of 12 feet ; while on a massive foundation the ground round which 

 is well protected by a building or trees the wave is slight. Deep under- 

 ground on a rock foundation with instruments such as I have had at my 

 disposal it is not perceptible, but it is not improbable that a residual effect 

 of the surface motion might be detected with more delicate apparatus. 

 The cause of the motion is not any immediate effect of temperature upon 

 the instruments, nor if we except the case where actual freezing of moisture 

 in the ground round and possibly inside one of the huts took place does it 

 appear to be due to expansion or contraction in or near the foundations 

 accompanying the acquisition or withdrawal of heat. 



The most active cause producing the movement which takes place 

 during the day may be the fact that the ground on different sides of an 

 instrument is unequally exposed to effects producing evaporation. The 

 retrograde motion during the night, which is smaller and more gentle than 

 that which has happened during the day, may be due to tlie unequal 

 condensation of moisture on two sides of a station. 



1. Effects accompanying Evaporation (Daylight Effect). — As the side 

 of a station from which most moisture is withdrawn to be dissipated in 

 the atmosphere has been relieved of a load, we should expect it to rise, 

 and this effect ought, vcv alluvium, to be perceptible to some depth. The 

 same area, because it is contracting like a drying sponge, may sink, but 

 tliis would be a superficial action. 



On open ground, under favourable circumstances, the load taken away 

 from a surface of earth by evaporation may amount to 4 or 5 lb. per 

 square yard, or from an area 20 yards square, about 1 ton. Experiment 

 has shown that 2 tons of wat^r taken out of a well and run off down a 

 hill will cause a pendulum at a distance of 20 or 30 yards to behave as if 

 the ground upon the well side had risen. If these premises are correct, 

 then an instrument well surrounded by trees or buildings, because the 

 evaporation is slight and is not likely to be much more marked upon one 

 side than it is upon another, should show but little motion. A pendulum 

 at a station freely and uniformly exposed upon all sides siiould also 

 show but little change. During the morning a north-south pendulum 

 would be expected to move slightly towards the west. For some hours 

 after the sun's meridian passage there would be a pause in such displace- 

 ment, after which a retrograde motion would set in. An instrument 

 with open ground upon its eastern side would, during the morning, be 

 expected to move westwards ; while at the same time another instrument, 

 M'ith the western side as an evaporation area, would move eastwai-ds. It 



