CHAPTER XV. BAROMETRICAL FLUCTUATIONS AND EARTHQUAKES—FLUCTUATIONS IN TEMPERATURE AND EARTHQUAKES. _Changes in the barometer and earthquakes._—Mallet, who collected together a number of examples of earthquakes which have occurred with a fall of the barometer, and a number which have happened with a rise, concludes that there are as many instances of the one as of the other. At the great earthquake of Calabria, in 1783, the barometer was very low. The earthquake of the Rhine (February 23, 1828) was preceded by a gradual fall of the barometer, which reached its lowest point upon that day. After the earthquake the barometer again rose. The earthquake of February 22, 1880, in Japan, was accompanied by exactly similar phenomena. Caldcleugh, who observed the heavy shocks in Chili (February 20, 1835), noticed that on February 17 and 18 the barometer fell 5/10 inches. Similar phenomena were observed before the succeeding smaller shocks. After the shocks the barometer again rose. Principal Dawson, speaking of the earthquakes of Canada, observes that some of the shocks have been accompanied with a low barometer. P. Merian, who examined the connection between the Swiss earthquakes and atmospheric pressure, found that out of twenty-two earthquakes observed in Basle between 1755 and 1836, thirteen of these were local shocks, of which eight were accompanied with sudden changes of pressure. Of the remaining nine, which were only felt slightly in Basle, no change in atmospheric pressure was observed. Of thirty-six earthquakes which, between 1826 and 1836, were felt in Switzerland, thirty were chiefly confined to Switzerland, and ten of these occurred with a low or falling barometer. Humboldt is of opinion that earthquakes only occur with changes in barometric pressure in those countries where earthquakes are few; and he gives examples where the regular variations of the barometer have gone on without interruption at the time of earthquakes. Frederick Hoffmann, who examined fifty-seven earthquakes which occurred at Palermo between 1788 and 1838, came to the following result:— The barometer was sinking in 20 cases „ „ rising in 16 „ „ „ at a minimum in 7 „ „ „ „ maximum in 3 „ „ „ undetermined in 11 „[113] The observations of M. S. di Rossi apparently show that the earthquakes in Italy chiefly occur with a barometrical depression and with sudden jumps in atmospheric pressure. Schmidt, who examined the earthquakes of the Orient, which occurred between 1858 and 1873, says that they were rare with a high barometer, but numerous when the barometer was low.