l899-'oo TRANSACTIONS . 1 31 



thoroughly satisfactory results. A great advance has been 

 made in getting instruments of the same pattern in various parts 

 of the world; but so far sufficient care has not been taken to have 

 them adjusted ahke, and hence the pendulums respond to vibra- 

 tions of different periods ; then it appears to me that the times of 

 the cut off of light are not reliable, and ^his of course is a source 

 of confusion. Matters, however, are improving in both respects, 

 and before long we shall have observations which are strictly 

 comparable, and results are likely to be more accordant. 



Professor Milne and several other observers have found a 

 diurnal wave in the records of seismograms — i. e. a slow tilting 

 which takes place in piers and buildings, especially on fine days, 

 for six or ten hours rapidly in one direction, and during the re- 

 mainder of the twenty four hours, more slowly and in an opposite 

 direction. This movement may be found underground where 

 changes in temperature are insignificant. The actual cause of 

 these movements is at present matter for speculation ; but the 

 theory which best explains the phenomena they present, as for 

 example that, on the opposite sides of a valley, it has been observ- 

 ed that movements take place simultaneously, but in an opposite 

 direction, is that these changes in the vertical are due to dif- 

 ferential changes on opposite sides of a station in the loads re- 

 moved during the day, or acquired during the night, by evapora- 

 tion and condensation of aqueous vapours. During the day it is 

 assumed that, by ordinary evaporation and the transpiration of 

 plants, the bottom of a valley loses more weight than its com- 

 paratively drier and less clothed sides. 



During a hot day the stream at the bottom of such a valley 

 should discharge fewer and fewer gallons of water, whilst the 

 valley bed, because it is relieved of load, should rise. For the 

 remaining fourteen or eighteen hours, because aqueous vapour is 

 condensed beneath the chilled surface of the ground, or as i 

 emerges from the ground on plant and other surfaces, the stream 

 in the bottom of the valley would increase its flow; and relatively 

 to the sides and bounding ridges of the va,lley, where we may 

 suppose the conditions for condensation to be less favourable, the 

 lower parts of the same would become heavier and consequently 

 sink. As to whether this concertina-like opening and shutting of 

 valleys, representing changes of slope of one or two inches in three 



