334 Professor John Milne [Feb. 12, 



wave. At Shide, in the Isle of Wight, where instruments like conical 

 pendulums are installed with their booms in the meridian on the 

 eastern side of a valley which runs north and south, the movements 

 are such that on fine days these booms point towards the sun, indicat- 

 ing that in the morning there is a downward tilting towards the east, 

 and in the afternoon towards the west ; at night the motion is east- 

 wards. The direction of this movement, which may have a range of 

 2" or 3", is, however, at the same time different at different places ; for 

 example, in Japan, on parallel ridges bounding a swampy valley, the 

 simultaneous movements on these ridges were found to be in contrary 

 directions ; that is to say, they were such that we may imagine the 

 trees on the opposite sides of the valley every day to ha ve performed 

 a slow bow to each other. 



Because these movements are practically confined to fine weather, 

 whilst in dull wet weather they are hardly discernible, we should 

 imagine them to be the result of expansions and contractions in the 

 surface soil, or warping of the piers carrying the instrument follow- 

 ing changes in temperature ; but when we find that they are practically 

 as marked in an underground chamber, where the changes in tempera- 

 ture are exceedingly small, the suggested explanation apparently fails. 



Another cause to which we may turn, as possibly throwing light 

 upon these movements, lies in the fact that, by the action of the sun, 

 there is on two sides of most observing stations a difference in the 

 load which, by evaporation, is carried up into the atmosphere 

 and there dissipated. As an illustration of this, if on one side 

 of an observatory we had a field of clover and on the other side 

 a surface of earth, the difference in the loads removed during a 

 day in summer would often exceed 12 lbs. per square yard. Be- 

 cause the clover side would be the one which would be the most 

 relieved, this would tend to rise, and the pendulum would swing 

 towards the uncovered surface. At night-time the causes leading 

 to a slow return of the pendulum towards its normal position would 

 be varied. For example, the area which during the day had lost the 

 most by evaporation would be the one presenting the greater number 

 of points for the condensation of moisture as it rose from the ground, 

 which, on the bare side, would be free to escape to the atmosphere ; 

 hence the clover-covered surface would, relatively to the ground on 

 the opposite side of the pendulum, grow heavy, be depressed, and the 

 pendulum take up a retrograde motion, which usually appears to be 

 somewhat less than the daylight displacement. 



Another phenomenon bearing upon the movement during the 

 night is the almost unstudied sub-surface precipitation of moisture. 

 Experiment has shown that in certain cases after sunset, when the 

 surface of bare earth is chilled or, in winter, frozen, aqueous vapour 

 rising upwards beneath such an area, instead of escaping to the atmo- 

 sphere is condensed underground, and the superficial soil grows 

 heavier. Soil which is tilled with stones probably shows this in 

 a marked manner ; each stone, being a good radiator, is at night 



