217 



have since been often observed by Mr. Dalton, 

 were considered as the forerunners of those 

 shocks, which were felt from 1748 to 1756. 



On the days when the earth is shaken by violent 

 shocks, the regularity of the horary variations o£ 

 the barometer is not disturbed under the tropics. 

 I have verified this observation at Cumana, at 

 Lima, and at Riobamba ; and it is so much the 

 more worthy of fixing the attention of natural 

 philosophers, as at St. Domingo, at the town 

 of Cape Francois, it is asserted, that a water 

 barometer * was observed to sink two inches and> 

 a half immediately before the earthquake of 

 1770. In the same manner it is related, that, 

 at the destruction of Oran, a druggist fled with 

 his family, because, observing accidentally, a few^ 



minutes before the earthquake, the height of the' 



r^tr • "iV-fifr ■ *»t ' ^fpn'r^r t ' "\ r \ * i * 'torr p'irirritO 

 sidered the oscillatory motion of the surface of the Globe as 

 the effect of an eleetric shock propagated from the air to the 

 earth. Ib. p. 642. Each of these theories admitted the 

 existence of a large black cloud' separating strata of air un- 

 equally charged with electricity, or 1 with nitrous vapours ; a'ntF 

 this cloud was seen at London at the moment of the first 

 shocks. I mention these reveries, to show to what errors w©% 

 expose ourselves in geology and physics, when, instead of 

 taking into view the whole of the phenomena that occur, we 

 suffer our attention to be arrested by accidental circumstances. 



* Currejolles, in the Journal de Phys., torn. liv. p. 106. 

 This depression answers only to two lines of mercury. The 

 barometer remained motionless at Pignerol, in April, 1808. 

 (Ibid. t. lxvii, p. 292.) 



