Dove on the Law of Storms. 317 



Further, if one of the phenomena be really an immediate conse- 

 quence of the other, we may not be able to infer with certainty 

 that the same effect might not have been produced in some oth- 

 er way. 



If barometric minima almost always occur when the atmos- 

 phere is agitated by tempests, on the other hand we frequently 

 see a very low barometer, when mild vernal breezes interrupt the 

 severe cold of winter and appear to introduce the temperature of 

 a more genial season. As it seemed however difficult to believe 

 that such gentle winds could cause any considerable disturbance 

 in the equilibrium of the atmosphere, the great diminution of 

 pressure on such occasions has been attributed to other causes. 

 The idea that the convulsions of the surface in earthquakes could 

 not be unconnected with the atmosphere, was one of such nat- 

 ural occurrence, that the barometer was always looked to in the 

 expectation of its indicating these phenomena at great distances. 

 This idea appeared to be confirmed, when, four days after the 

 destruction of Messina in 1783, the barometer in Europe fell unu- 

 sually low. Van Swinden accordingly inferred a connection be- 

 tween the two phenomena ; but on a comparison of the meteoro- 

 logical observations made at the time, and recorded in the Man- 

 heim Ephemerides, Brande found that on the 9th of February the 

 barometer fell below its average height by 14 lines in Lyndon, in 

 Rutlandshire ; 13 J in Amsterdam and Franeker ; 12| in Dun- 

 kirk ; 12J in Middleburg; 12| in Paris; 11£ in Laon, Nantes, 

 and Cambray ; 10J in Brussels, Chartres, Poictiers, and Rochelle ; 

 10 in Troyes and Montmorenci ; 9 at Gottingen, Mayence, Metz, 

 Limoges, and Bordeaux ; 8 at Copenhagen, Erfurt, Wurzburg, 

 Lyons, Mezier in Guyenne, and Oleron ; 7 at Spydberga in Nor- 

 way, Stockholm. Berlin, Vienna, Manheim, Geneva, and Vienne; 

 6 at Sagan, Prague, Regensburg, on the St. Gothard, and at 

 Montpelier ; 5 at Marseilles and Montlouis ; 4 at Ofen and Padua ; 

 3 at Petersburg, Mafra, Bologna, and Rome. Thus it appeared 

 that the barometer was lowest in England and Holland, and that 

 in approaching Italy it differed less and less from its mean height, 

 so that the independence of the two phenomena became highly 

 probable. 



If, as in this instance, such simultaneous observations some- 

 times serve to show, that what had been regarded as evidence of 

 essential connexion was merely an accidental coincidence be- 



