72 On the Whirling Action of the 



4. It has been noticed, also, that the directions given to bro- 

 ken limbs and other bodies, by the successive changes in the di- 

 rection of the wind as the tornado passed over, have been found 

 in opposite courses of change, on the two opposite sides of the 

 track. 



This fact, too, has been strongly urged as disproving a rotary 

 motion. But, unfortunately for the objection, this effect accords 

 fully with the rotary action of a progressive mass of atmosphere ; 

 as is well known to all who clearly understand the theory of ro- 

 tary storms. 



In all such whirling masses the successive changes in the di- 

 rection of the wind, result solely from their progressive motion, 

 and necessarily take place in opposite directions or courses of 

 change on the two opposite sides of the advancing axis. This 

 indication fails, therefore, as a theoretic test ; and I now proceed 

 to notice others, which are peculiar to a progressive whirhng ac- 

 tion. 



5. In considering further the effects of such action, we may 

 expect to find that the greatly increased activity of gyration 

 which is always observed near the centre of a vortex, will be 

 indicated by a more violent and irregular action in and near the 

 path pursued by the axis of the whirlwind, than is found under 

 its more outward portions. 



This effect is often strikingly exhibited in the path of torna- 

 does ; while, in the supposed ascent of a non-whirling column, 

 it would seem that no part of the surface would be so much ex- 

 empted from its action, and particularly from its power of prostra- 

 tion, as that lying near its centre. 



6. As the effect of rotation must be to produce, on one side 

 of the advancing axis, a reverse motion which is contrary to the 

 course of the tornado, it is evident that on this side the prostrating 

 power will be much lessened ; that the cases of prostration, there- 

 fore, will be here less numerous ; and that some of these, at least, 

 will be produced in a backward direction, more or less opposite 

 to the course of the tornado. By this criterion, not only the 

 whirling movement, but the direction of the rotation also, may 

 be clearly ascertained. 



This effect is best observed by comparing the two opposite 

 margins of the track, and is strongly exemplified in the case be- 

 fore us. Here we find, that most of the trees prostrated within 



i 



