382 The Great Wiltshire Storm. 



were both blown inwards, sbows that the air must hare been 

 rarefied by the passing of the cloud, and that they were driven 

 in by lateral pressure. But not to prolong the question unneces- 

 sarily, I come now to what I apprehend to be the greatest feat of 

 the storm, which was the breaking off and hurling to a distance 

 of nearly 60 yards before they struck upon the soft ground, the 

 heavy tops of three elm trees, standing just above Cherhill Mill, 

 whose length was about 25 feet, and whose weight may be conjec- 

 tured from the fact, that Mr. Beynolds assures me it required three 

 horses, and even then was as much as they could do, to drag them 

 one by one into his yard. The above theory of the production of 

 a vacuum and its absorbing tendency, aided by the force of the 

 gale from without, accounts very satisfactorily to my mind (and 

 nothing else will account) for this extraordinary feat ; as well as 

 for similar instances, of which there are several, of other trees and 

 other large limbs hurled a considerable distance ; among which I 

 would particularise one at Quemerford Mill, another at Mr. Maun- 

 drell's farm blown across an entire meadow, and another in Barrow- 

 way at Tatesbury. The removal of the three entire roofs, viz. of 

 the cottage at Cherhill (measuring 16 by 13 feet) of the shed at 

 Yatesbury (41 by 15 feet) and the cattle shed at Monkton (53 by 

 16 feet) though to be accounted for on the same principle, differs in 

 certain respects. In neither case does it appear that the walls 

 supporting those roofs are in any degree injured, but the roofs seem 

 to have been lifted up by some strong upheaving force, as the 

 cloud passed over, and then a current in the direction of the storm 

 carried them on. And this (I learn from Mr. Rowell) is no un- 

 common occurrence during tornadoes, for (I quote again from his 

 book) " the great diminution of atmospheric pressure within the 

 whirl is shown by the fact that in violent tornadoes, the windows, 

 doors, &c, of buildings near the centre of the line of the tornado, 

 are very often burst outwards, as if from the expansion of the air 

 within the building on the sudden cessation of external pressure : 

 even the cellar floors of buildings have been burst upwards during 

 such storms, where it has been impossible for the wind to get 

 beneath them to force them up." (276). The same principle of 



