prevalent Disorders, fyc.> with Volcanic Emanations, 157 



on Professor Olmsted's Theory.'' (A. «/., xxvi. 320.) In 

 reference to this, I have before (VII. 387.) stated an opinion 

 respecting a periodical derangement of the earth, as in- 

 fluenced by a general law, and producing corresponding at- 

 mospherical and meteoric phenomena. This may be a bold 

 assertion ; but there are coincidences of so strong a character 

 to support it, that I have sent it out unqualified. 



It is observable, though not often observed, that there are 

 particular months and particular days of the month, on which 

 certain meteoric phenomena have frequently occurred. A 

 few examples shall suffice ; and as the chief topic of the pre- 

 sent paper is connected with the wind, I will choose the subject 

 of hurricanes to illustrate the position. 



August 31. 1833, and August 31. 1816 (at intervals of 

 seventeen years), were the dates of extraordinary hurricanes 

 in Europe. Now, within a day or two of this date, on Sept. 2. 

 and 3. (the night before Cromwell died) in 1658, there was 

 an awful hurricane all over France, Italy, the Mediterranean, 

 Germany, and Great Britain. There are also many instances 

 of hurricanes occurring about a particular day in May and 

 September, as the hurricanes at Calcutta and New Orleans on 

 May 21. 1833, and at Arracan, May 14. 1834; the hur- 

 ricane at Dominica on Sept. 15. 1816, and Sept. 20. 1833. 



A writer (J. M. [Mr. Main]) in this Magazine (I. 180.) 

 has stated that the wind near London always blows from the 

 eastward at the beginning of May, there being only two 

 exceptions in the space of twenty-two years. His observa- 

 tions tend to strengthen my idea of the periodicity of atmo- 

 spherical phenomena. 



The last few days of October have also been often celebrated 

 by the occurrence of remarkable hurricanes in different parts 

 of the earth. The hurricane of Balasore, on Oct. 31. 1831, 

 of which the particulars are given above (p. 149.), is, perhaps, 

 as remarkable an instance, as any on record, of violence and 

 devastation. On Oct. 31. 1753, occurred a hurricane at 

 Wandiwah, about sixty miles from Madras ; a distance which 

 seems to have been the radius of the storm. (Orme.) On 

 Oct. 29. 1768 occurred a destructive hurricane off the coast 

 of Coromandel. On Oct. 30. 1731 happened a violent 

 tornado in Dorsetshire (P. T.); and, on Oct. 30. 1669, a 

 formidable hurricane in Northamptonshire.* 



* October affords many other instances of violent hurricanes and 

 whirlwinds : as, Oct. 13. 1670, at Braybrook, Northants ; Oct. 6. 1756, at 

 Wigton, Cumberland ; Oct. 20. 1763, at Madras ; Oct. 2. 1746, at Madras ; 

 all Oct. 1833, in Europe and America ; Oct. 23. and 30. 1834, in Britain, 

 Holland, Germany, France, and Mexico. 



