and 071 Shooting Stars i?i general. 225 



meteor was seen by Messrs J. and L. Chandler, to move nearly 

 parallel with the horizon, at an elevation of three degrees, accord- 

 ing to one, and of four degrees, according to the other. It was 

 seen in the N. E. quarter, but the precise direction could not be 

 given, as the estimate was made when they had no opportunity 

 of returning to the spot of observation. The preceding obser- 

 vations are the most precise of any I have been able to obtain. 

 They were all made by the aid of instruments, and the chief 

 error therefore to be apprehended, is that arising from the diffi- 

 culty of exactly remembering the apparent position of the me- 

 teor. It is believed, however, that in the above observations, this 

 error is small. In this first comparison, I neglect entirely such 

 observations as give mere estimates of elevation by the eye ; for 

 it is a remarkable fact, that almost every one over-estimates an- 

 gular elevation near the horizon. I have made the computation 

 from the above data, and find the perpendicular elevation of be- 

 ginning 28.5 miles. The place where it then stood in the ze- 

 nith was in Lycoming county, Penn,, Lat. 41° 16', Long. 1^ 

 west from Washington. At its explosion, its height was 32.1 

 miles, and the place where it then stood in the zenith was in 

 Upper Canada, Lat. 44° 7', Long. 2° west. At an intermediate 

 point, its height was 34.8 miles, and it was then vertical over 

 Monroe county, N. Y., Lat. 43° 0', Long. 0° 46' west. The 

 length of path then was 218 miles, and its mean course N. 13 J° 

 W., passing vertically over Rochester in the State of New York. 

 In this computation I have aim^d to lake the positive equal to 

 the negative errors, and the sun. ji tue errors, disregarding their 

 signs, a minimum. The beginnuig, as I have here assigned it, 

 rests upon the observations at Clinton, Hudson, and Aurora. It 

 appears not to have been seen so early in its course either at 

 Hanover, Buffalo, or Ann Arbor. The middle point of its path, 

 near Rochester, appears to have been observed at all the stations. 

 The observations at Hudson and Aurora having been made near 

 each other, I consider as one observation, and take their mean. 

 It is impossible perfectly to satisfy the observations. The result 

 I have above given, makes the errors of the observations as fol- 

 lows : Hanover + 3° 45' ; Chnton - 2° 54' ; Buffalo - 2' ; Hud- 

 son - 5' ; Ann Arbor — 44'. The positive are equal to the nega- 

 tive errors, and I am unable to assign the meteor a position which 

 shall diminish the sum of the errors. The errors at Buffalo, Hud- 

 VoL. XXXY.— No. 2. 29 



