INTRODUCTION. XXVii 



the barometer falls successively from north of west to south of 

 east, 1 it seems highly probably that these storms of oblong form 

 move towards the south of east (passim.*) 



In the West Indies, from Barbadoes to Jamaica, it is known, by 

 the invaluable labors of Redfield and Reid, that the hurricanes 



1 See diagram at the end of Appendix, pp. 550 and 551. 



2 It seems probable, from the following facts, extracted from Orlando Whis- 

 tlecraft's Climate of England, that thunder storms in England also travel from 

 north of west to south of east. 



Page 2. In summer, too, a south east wind always prevails in the eastern 

 countries before the great thunder storms, which by night spread themselves 

 over the greatest part of England. These storms are evidently aided in their 

 rise and progress, or passage to the eastward, by a south west or west wind 

 behind them, which combats against the sea breezes so usual in Kent, Essex, 

 Suffolk, &c., from south east or east during hot days. 



Page 7. The south east wind blows in July with the hottest and clearest 

 days for nearly a week in the eastern counties, until vapors arise with a 

 south west current and form distant ranges of rocky clouds in the south west 

 horizon. These and the white and round detachments of cirrocumulus in 

 trains across the zenith, are the first clouds after clear and hot days indicating 

 a change, which takes place by the south western clouds coming on with a 

 severe thunder storm by night for many hours, while the wind continues 

 east or south east, until it passes, when it veers to the south west, whence the 

 storm arose. 



Page 10. On the night of the 9th August, 1787, a dreadful thunder storm 

 came on. This was introduced by a fine and hot week, with easterly wind, 

 and the storm came on as usual from south west, whither the wind afterwards 

 veered. 



Page 22. In a hot and clear sunshine, so much moisture is drawn up by 

 evaporation, that many distinct cumuli are formed by 10, A. M. in the other- 

 wise clear sky. These increase in size till about 2, P. M., after which they 

 decrease, and at sunset we have none again in view. 



Page 27. The cumulus is truly the day-cloud, beginning to form itself in a 

 previously clear morning, increasing till 2, P. M., and then decreasing, until 

 at sunset no cloud again appears. In this case we see a mere speck of vapor, 

 at about 10, A. M. in a summer's day, accumulating till a semicircular body 

 is formed, having a flat base, while the upper part, is somewhat rocky in ap- 

 pearance. 



Page 28. As the cumulus is about to pass into the nimbus, the middle of 

 the cloud will represent the neck of a mushroom, and the summit spreads and 

 overhangs the base in a most striking manner, and the tops of these clouds 

 may be compared to the ebullition or at least to the effervescing of some fluid. 



Page 30. It has been long observed by meteorologists, that a south east 

 wind (in the eastern counties at least,) precedes the most violent thunder 

 storms, and that the storm itself works its way in a higher current from the 

 westward. 



Page 67. On the 23d May, 1830, an extraordinary cloud began to rise at 

 6, P. M., and veiled the south west in blackest hue for two hours. It made 

 its way against the opposing current below with a continued blaze of blue 

 lightning, and continued over us about an hour. 



And on the 3d of June, at 4, P. M., a violent storm came up against the 

 lower current, which was east north east. 



On 2Gth June, 1H33, a thunder storm came from south west, wind north 

 east. The same phenomenon occurred on the 24th August, 1834, and also on 

 the 13th October, 1830, and 29th June, 1838. 



