199 [ 23 ] 



9lh. The air thus rushing in under the cloud on ail sides will ascend, 

 and carry up its vapor with it, and will condense it into cloud by the cold 

 of expansion from diminished pressure, as before, and thus the process of 

 cloud-forming will be continued. 



10th. As the principal part of this upmoving current of air is in the upper 

 regions of the atmosphere, which move all the year, over the United States, 

 from west to east, the cloud, with its whole column of light air, must also 

 move in the same direction. 



" Dr. Hare has turned his attention to the electrical phenomena which 

 accompanies the more violent exhibitions of the storm when it assumes 

 the form of a tornado. He agrees with Mr. Espy in opinion as to in- 

 ward and upv\^ard direction of the wind towards the middle of the storm, 

 but differs from him with respect to the cause of the current. He proposes 

 a definite theory of the phenomena of the tornado, founded on well known 

 laws of electrical action, combined with facts of observation. It has been 

 fuUy^ established by experiment in different parts of the earth that there is 

 an accumulation of positive electricity in the upper regions of the atmo- 

 sphere, that the surface of the earth relative to the lower stratum of air is 

 slightly negative, and that space void of air may be considered an electrical 

 conductor. From these facts Dr. Hare infers that the surface of the 

 earth, the surrounding atmosphere, and the space immediately around 

 and exterior to the atmosphere, form three concentric spheres, of which the 

 outer and inner are constantly charged with opposite electricity. The 

 arrangement is, therefore, precisely that of a charged leyden jar, of which 

 the exterior sphere is analogous to the outer coating, the surface of the earth 

 to the inner, and the intervening atmosphere to the non-conducting glass. 

 The clouds are insulated conductors floating between the two coatings, and 

 therefore liable to be variously charged by induction as their position rela- 

 tive to each other and the two coatings is varied. In some cases the ar- 

 rangement may be such as to form a series of steps of intermediate con- 

 ductors, each electrified by induction, so as to produce a violent ascending 

 current ox conveetioe discharge from the surface of the earth upwards. 



"■ The theories of Redfield, Espy, and Hare may not be found incompati- 

 ble in all points; and whatever may be the future state of our knowledge 

 relative to them, they point to definite objects of inquiry, which cannot 

 fail to reward proper observation with a rich harvest of important results." 



The idea occurred to me that more might be learned liom a complete 

 analysis of a single storm, than from a partial analysis of several storms; 

 and that any storm of strongly marked characteristics, if fully investi- 

 gated, mast p'ove a complete experimeMum crucis, at least between 

 the theories of Messrs. Redfield and Espy. I accordingly selected the 

 storm of December 20, 1836, for a thorough examination. I obtained 

 barometric observations fiom 27 different stations within the United States 

 and the neighboring British possessions. I also obtained meteorological 

 journals, not containing barometric observations, from 28 military posts, 

 from 42 academies in the State of New York, and from five other stauons; 

 making 102 in. all, besides several stations beyond the probable hmils of 

 the storm. . 



In analyzing these materials, the barometric observations were all graphi- 

 cally represented by curves, showmg the fall and rise of the barometer, 

 with the time of its minimum height. Then joining by a hnc all those 

 places where ih« minimum of the barometer occurred at the same instant, 



