412 METEOROLOGICAL CORRESPONDENCE 



Remarks. — Thimder storms are most frequent and violent in tropical regions ; 

 they seldom occur in high latitudes, particularly on the ocean. Tliis is proba- 

 bly due to the coldness of the surface and the consequent small amount of vapor 

 in the atmosphere, a thunder storm, being as we have said in the foregoing 

 remarks on the Gulf Stream, a large ascending column of air rendered of less 

 density than the stratum immediately above, by an excess of heat and moisture. 

 Although the difference of density in this sti-atum may not be sufficient of it- 

 self to produce the upward motion at the level of the sea, yet the transfer of the 

 same stratum up the side of a mountain would produce a precipitation of mois- 

 ture, a liberation of latent heat, a vertical conducting column, and, in short, all 

 the conditions necessary to a thunder storm. Hence the more frequent occur- 

 rence of this meteor in mountainous regions in the same latitude than on the 

 fthores of the ocean. 



The statements of Mr. Bannister, as to the position and sound of the aurora, 

 are in accordance with the testimony of some of the northern voyagers, who re- 

 port haviug seen the beams projected on clouds and also against a hill at the 

 distance of a few thousand yards. Indeed, it would appear, from an attentive 

 consideration of all the facts relative to the aurora, that it consists of beams or 

 discharges of electricity between the earth and the upper regions of the atmo- 

 sphere, j)arallel to the dipping needle at the place of observation. These dis- 

 charges, according to the theory, must, when a corona is formed, take place 

 immediately around the observer, since in this case the corona is simply the van- 

 ishing perspective point of the parallel beams. We know that the appearance 

 of the aurora at so small an altitude, and also the occasional emission of sound 

 by the meteor, have been doubted, but Ave have never been able to infer from a 

 priori consideratious why the statement in regard to both of these phenomena 

 may not be true. Late investigations leave no doubt as to the electrical nature 

 of the aurora, and as the visibility of the discharges of this agent must depend 

 upon the greater or less degree of the conductibility of the atmosphere, they may, 

 in some cases, probably in dry weather, appear luminous down to the earth. The 

 character of the sound is also similar to that which is produced by electricity. 



In regard to the cold, it is a general observation that the thermometer sinks to 

 the lowest degree during a perfect calm with clear sky. At this time the radia- 

 tion from the earth is most intense, since all the local vapor is precipitated, and 

 scarcely any obstruction is offered to the free passage of the rays of heat into 

 celestial space. In this condition in a high latitude a wind which blows for some 

 time, from any direction, is accompanied by a rise of temperature, since it must 

 finally come, even across the pole, from a region of more moisture, though the 

 water which it bears may be in a frozen condition, and thus produce a mistiness 

 which would check radiation. Besides this, the wind from all points of more than 

 one-half of the horizon, must come directly from a more southern quarter, and if 

 it continues for a sufficient length of time, must needs have a higher tempera- 

 ture, from whatever direction it may come. Precisely at the poles all winds must 

 be from the south. 



The late Mr. Kennicott described to me a phenomenon which has also been 

 mentioned by Sir George Simpson and other northern voyageurs ; it consists of 

 a rustling or crackling noise produced by the breath, when the temperature of 

 the surrounding air is below — 40°. He described it as prodvicing a startling 

 effect when first heard, the sound appearing to come from a point either to the 

 right or left of the observer when in motion. 



This effect is probably due to the sudden collapse of the vapor of the satu- 

 rated air as ft escapes from the lungs, and the consequent sudden reduction of 

 the volume. That the condensation 'of the vapor may be sufficiently rapid in 

 this case to produce a sound is scarcely surprising, when we consider the rapidity 

 with which steam is condensed by water of the ordinary temperature in the opera- 

 tions of the steam engine. 



