388 Tornado in N. E. part of Ohio, Feb. 4, 1842. 



bearings would begin from the west, and end northeast or east. 

 Still farther they would begin southwest, and end east or south- 

 east ; and upon the extreme margin the wind would blow to- 

 wards southeast. All this agrees remarkably well with observa- 

 tion. The house and barn H lay strewed between the directions 

 N. 88° E., and S. 50° E. 



We may now perhaps explain some of the anomalies noticed 

 in the bearings on page 285. The anomaly in group third may 

 perhaps be due to a more easterly motion of the whirl than at the 

 fall of the fourth tree, or to the fact that being nearest the vortex 

 the proper motion of the whirl was greatest, and the bearing of 

 a tree consequently less influenced by the progressive motion of 

 the tornado. This last cause might perhaps explain the anoma- 

 lies in groups sixth and eighth. The anomaly in group four- 

 teenth may perhaps require us to admit a slightly undulatory 

 motion of the vortex, such as is frequently seen in small whirl- 

 winds. In group tenth also, although the rotation is uninter- 

 rupted, the variety of directions seems most naturally explained 

 by such a supposition. I think it moreover not improbable that 

 the direction of the wind was materially different at different ele- 

 vations above the earth's surface upon the same vertical line, so 

 that the bearing of a prostrate tree might be influenced by its 

 height. 



We also observed the bearings of a good many fallen trees in 

 the vicinity A A, B B, but the trees were here more sparse. Near 

 A A, the bearings were all comprised between N. and N. 28° E. 

 Near B B, the directions were about the same as at O ; but 

 from their distance the trees seldom lay upon each other. Near 

 the northwest border were measured N. 84° E., N. 88° E., N. 

 89° E., East, S. 82° E. In the vicinity F F the following ob- 

 servations were made. 



These observations are similar to those made near O O, and 

 present two similar anomalies, which I have marked with an as- 



