prevalent Disorders, tyc, with Volcanic Emanations. 135 



(see VI. 299. note) who consider the sun to be the cause of 

 earthquakes. The position I defend suffers nothing, there- 

 fore, on that account. The convulsions of the air were very 

 extensive at that time for several days. On Sept. 24., the town 

 of Materna, and the environs of Basilicata (Naples), were 

 visited by a tremendous hailstorm ; the stones were as large 

 as walnuts, and lay a palm (10^ in. deep); 14,000 panes 

 of glass were broken. At Monte Video, and throughout 

 the Rio de la Plata (34° 30' s., 56° w.), there were de- 

 structive and tremendous gales on Sept. 25, 26, and 27. 

 On Sept. 25., also, Smyrna was visited by a furious tornado, 

 which travelled northwards to the Black Sea, where a hur- 

 ricane blew from the south, contrary to the usual wind 

 there, on Sept. 28. If these gales were actually connected 

 with the convulsions of the air in the West Indies, as they 

 probably were, the interval of time corresponds with the dis- 

 tance, at the rate of under thirty miles per hour for the aerial 

 impulse. But whether this calculation be, or be not, pre- 

 posterous, I see no reason to question a connection of all 

 these phenomena with each other. The hurricane of Au- 

 gust 12. 1830 had a progress, in an ascertained direction, of 

 about twenty miles an hour; while that of Aug. 10. 1831 

 travelled at the rate of fifteen miles an hour: but, in the 

 atmosphere itself, unimpeded by friction with the earth, and 

 flowing along in a right line, not in a circular sweep like the 

 hurricane, the air must rush in to supply the vacuum caused 

 by even a distant hurricane, with a speed which baffles cal- 

 culation. Nor is the rate of 126 miles per hour across the 

 Atlantic, on Sept. 20., a rate at all exaggerated, if we con- 

 sider the tornado and hurricane as indices of an aerial tor- 

 rent. There is one anomaly in the hurricane of September, 

 if what is said above (VIII. 22.) be of any value; it did not 

 occur at a change of the moon. That luminary was full on 

 Sept. 17., the first and last quarters occurring on the 10th 

 and 26th.* May not this give strength to the notion that it 



* It has occurred to me, in the course of my reading on this subject, 

 that earthquakes, volcanic eruptions, hurricanes, &c, occur frequently either 

 at the time of the lunar changes, or, at the most, three or four days after- 

 wards. Such was the case here; in the inundation at Petersburg in 1777; 

 in the eruption of Vesuvius on Aug. 27. 1834, the day of the last quarter, 

 &c. ; and, as I understand from my friend Mr. Lloyd, Dr. Foster has ob- 

 served the same things, especially respecting the new or full moon, of all 

 great eruptions of Vesuvius, Etna, &c. &c. Should this be the case, I 

 shall gain thereby another link for my argument; nor do I see any mate- 

 rial objection to the supposition ; for, if the moon attracts the sea, why may 

 it not also attract the earth, and the volcanic fluid thereof? Dr. Foster, in 



k 4 



