330 EARTHQUAKES. 



Many of these changes may be due to changes in 

 temperature, variations in moisture, and other local 

 actions. Some of them, however, are hardly explicable 

 on such assumptions. The fact that the general direction 

 in change of the vertical, as indicated by a tromometer 

 standing on the same column with the levels, showed 

 that the change which was taking place was rather in the 

 column than in the instruments. 



The fact also that at the time of a barometrical 

 depression a pulse-like surge can be seen in the levels, 

 having a period averaging about three seconds and some- 

 times amounting to about one second of arc, is a pheno- 

 menon hardly to be attributed to sudden fluctuations in 

 moisture or temperature, but indicates real changes in 

 level. ^ 



In addition to variation in the bubbles of levels 

 which come on more or less gradually, we have many 

 recorded instances of sudden alterations taking place in 

 these instruments. 



Examples of what may have been a slow oscillating 

 motion of the earth's crust are referred to by Mr. George 

 Darwin in a Eeport to the British Association in 1882. 



One of them was made by M. Magnus Nyren, at 

 Pulkova, who, when engaged in levelling the axis of a 

 telescope, observed spontaneous oscillation in the bulb of 

 the level. 



This was on May 10 (April 28), 1877. The complete 

 period was about 20 seconds, the amplitude being 1'5'^ 

 and 2''. One hour and fourteen minutes before this he 

 observes that there had been a severe earthquake at 

 Iquique, the distance to which in a straight line was 

 10,600 kilometres, and on an arc of a great circle, 12,500 

 kilometres. On September 20 (8), in 1867, Mr. Wagner 



' See * Earth Tremors,' p. 309, experiments of M. d'Abbadie, &;c. 



