336 Supposed Connection of Meteoric Phenomena, 



and the prevalence of aurorse on the nights of these meteors 

 (see M. N. H., vii. 291. 294. 387.; and Journal of Franklin 

 Inst, xv. 90.)]. Evidence is adduced by Mr. Espy to show 

 that, at Fort Gibson, the display was intermittent, alternating 

 light and darkness at intervals of about thirty or forty seconds ; 

 and that, " at one or two of the military posts far west, all the 

 meteors moved horizontally." (p. 91.) The radiant there 

 was also in the south-eastern horizon. How, it is asked, does 

 this "agree with Mr. Olmsted's?" (p. 92.) 



Having thus upset Olmsted's theory, Mr. Espy proceeds to 

 introduce his own. He begins by asking the following ques- 

 tions : — • 



" 1. Do auroral arches always move towards some southern 

 point ? 



" 2. Do the streamers always lean in the direction of the 

 motion of the arch ? 



"3. Is there always a sudden depression or elevation of 

 the dew point about the time of a brilliant aurora?" (p. 159.) 



The first two queries, he says, may be answered in the af- 

 firmative, so far as his own and others' observations tend. 



" 4. Are clouds or haze in our [i. e. Pennsylvania] latitude 

 generally formed behind the arch?" Facts, he says, have 

 come to his knowledge to establish the affirmative of this ; 

 and it may account, if so, why in England and in America the 

 auroras are oftener seen in the north than in the south. He 

 quotes the brilliant aurora of May 17. 1833, as an example. 

 The wind had been north ; but the lower current changed to 

 to the s. w. below, whilst the upper current continued from 

 the north. 



" 5. Are there also two currents in different directions, at 

 the time of auroras; and is the upper one generally from the 

 north ? Or, where the upper and lower currents are in the 

 same direction, from the north, does the upper move fastest?" 

 (p. 160.) 



He thinks this will, some day, be answered in the affirm- 

 ative. " Besides, many meteorological phenomena can easily 

 be accounted for, on the supposition of an upper current of 

 air from the north, at the time of their occurrence, even when 

 this current is not indicated by clouds." Mr. Espy considers 

 the sudden depression of the dew point as indicative of a north 

 current at the time of its occurrence. The dew point on 

 Nov. 12. 1833, was, at 1 o'clock, at 57° F. ; at 6 p. m., at 40°. 

 It rained a little during the day ; the wind was s. w. ; the 

 evening was perfectly clear. An aurora was expected: none 

 was visible. On the 12th, there were no clouds in the upper 

 stratum ; on the 1 1th, the lower air was from the east, and the 



