AURORA BOREALIS. 

 Table II. 



219 



If we combine the precediug observations in Table II, we shall find that the 

 lower limit of the auroral light was elevated forty-six miles above the earth's 

 surface, and that its southern margin was vertical over the parallel of 38° 50' 

 north latitude in Virginia. 



Now it is considered as established that the auroral streamers are luminous 

 beams sensibly parallel to the direction of the dipping needle. But the dip of 

 the needle in latitude 38° 50' in Virginia is 71° 20' ; and if we draw a line D, 

 making an angle of 71° 20', with the curve line A B, which represents a por- 

 tion of the earth's surface, we may assume that the line C D represents the 

 Bouthern boundary of the auroral illumination. If, then, we assume that the ob- 





servations of Table I were made upon the point D, we shall find that the upper 

 limit of the auroral light was elevated 534 miles above the earth's surface, and 

 that its southern margin was vertical over the parallel of 30*^ 40' north latitude 

 ill Virginia. 



The following table presents a summary of the most definite observations of 

 the aurora of September 2, 1859, made generally about 2 a. m., Havana time- 



