The Observation of Earthquakes. 159 



at a depth of 19 feet in sandstone. On the opposite sides of two 

 valleys where instruments have been established, it is found that 

 these daily movements take place simultaneously, but in opposite 

 directions. During the day we may picture a valley opening, and at 

 night as closing. 



Bearing in mind that these diurnal movements are only marked 

 on bright, sunny days, and are practically absent in dull, wet 

 weather, we may seek for their explanation either as a general expan- 

 sion of the ground under the influence of sun heat, or to a diurnal 

 loading and unloading of a valley bottom relatively to that which 

 takes place upon its sides. 



The fact that these deflections of the pendulum may be observed 

 in cellars and chambers excavated in rock where temperature is 

 practically constant, precludes the idea that they are due to any local 

 heating of the foundation on which the instrument may be installed. 

 It, however, does not preclude the idea that there may be a general 

 superficial warping of a district as an effect of solar radiation. It is, 

 however, remarkable that this action should extend to the depths at 

 which it has been observed. 



That valley beds convey more load at night than thev do 

 during the day is suggested by various observations. Engineers have 

 shewn that under normal conditions certain streams carry the most 

 water at night time. This is also true of certain drains and wells. 

 The causes leading to these conditions may be various. A nocturnal 

 increase in the flow of sub-surface water may be attributed to the 

 expansion of air in soil by the slowly-descending heat of the previous 

 day, which forces interstriaal water into channels of easiest flow. 



Another explanation rests on the fact that during the day 

 evaporation and vegetable transpiration are at a maximum, whereby 

 the flow of sub-surface water is diminished. 



At night, with the cessation of these activities, the flow is 

 relatively increased, and valleys receive their greatest load, with the 

 result that their sides close inwards. To support the idea that water 

 load plays an important part in the fluctuation of level we have 

 the repeated observations that during wet weather, when we see 

 water accumulating in the beds of valleys, the heeling over of the 

 booms of horizontal pendulums is towards the loaded district. 



The main point, however, to which I desire to draw attention 

 is not so much the explanation of curious phenomena, but simply to the 

 fact that Horizontal Pendulums may, under certain conditions, be 

 influenced by hidden water loads. The investigations that this 

 suggests should be of particular interest in certain parts of South 

 Africa. For example, do the fluctuations of the subterranean water 

 tapped in the Karoo affect the surface level ? If they do, then the 

 revelations of a pendulum may play an important part in the opening 

 up and the settlement of a district. 



Whether this type of seismograph will be used as an assistance 

 in the prediction of weather is a matter worthy of some consideration. 

 The Barometer gives the atmospheric pressure where it is installed. 

 The Horizontal Pendulum, under certain conditions, swings to the 



