New Brunswick in New Jersei/, U. S.,Jime 19, 1835. 25 



direction on this side being S. 47° West. We have thus a 

 mean difference in the inclination of the fallen trees on the 

 two exterior portions of the track, of no less than eio-hty- 

 eight degrees. These indications seem conclusive also in 

 favour of the whirling action in the direction from right to left. 



10. Although of less importance, it should be mentioned 

 that the diminished action of the tornado which is commonly 

 observed on the hill sides and summits over vi'hich it passes, 

 and the greatly increased action in the bottoms of the valleys 

 and even in deep ravines, afford a strong argument against 

 ascribing the effects to the ascent of a non-whirling rarefied 

 column ; as the latter, it would seem, must act with greater 

 force on the hill sides and summits than in the bottoms of 

 valleys. The general correctness of the observation here 

 stated cannot justly be questioned. 



11. The sudden and extraordinary diminution of the at- 

 mospheric pressure which is said to take place at the points 

 successively passed over by the tornado, causing the doors and 

 windows of buildings to burst outwards, seems to afford 

 strong confirmation of a violent whirling motion, for an effect 

 of this kind is necessarily due to the centrifugal and upward 

 force of the vorticular action. There are no other means 

 known by which such an abstraction of pressure can be ef- 

 fected in the open air. An increase of calorific elasticity, if 

 such were produced, either generally or locally, would not 

 greatly disturb the equilibrium of pressure, being resisted by 

 the surrounding and incumbent weight of the entire atmo- 

 sphere. Besides, the immediate effect of such increased elas- 

 ticity might rather be to burst imoard the windows and doors 

 of buildings exposed to its action. 



Some of the more important indications mentioned above, 

 appear also from Prof. Bache's observations; although the 

 latter are not definitely located by him, as regards the ex- 

 treme borders of the track. Thus, in figure 7 of Prof. 

 Bache's paper, assuming the course of the tornado to be east, 

 and rejecting a few observations near the centre, to avoid 

 error, we find in twenty observations on the right side of the 

 track, a mean inward inclination of sixty-four degrees, and 

 for nine observations on the left side, a mean inclination, 

 reckoned inward and backward from the course, of one hun- 

 dred and four degrees, being fointeen degrees backward. 



It is stated by Prof Bache, " that the trees lying perpen- 

 dicular to the track of the storm, are not those furthest from 

 the centre of that track." This generalization accords with 

 my own observations; but can hardly be reconciled with an 

 inward non-whirling motion in the tornado. 



It may appear to some that in the case of a whirlwind the 



