150 PROCEEDINGS OF THE AMERICAN ACADEMY 



are the third parabolic elements computed for this comet, (originally 

 published by S. C. Walker, Esq., in the National Intelligencer of 

 April 26th,) from Cambridge observations, April 11th, 14th, and 19th, 

 and Bessel's elements for the comet of 1748. 



1848. Mean Berlin Time. 1748. Mean Paris Time. 



T June 8^-.23220 T 18^-.89401 



9> 30° 32' 7" _ 9> 33° 8' 29" 



i 66 55 12 i 67 3 28 



n 267 13 6 TT 278 47 10 



q 0.892703 q 0.625357 



Motion direct. Motion direct. 



" The discrepancies between these two orbits are not greater than 

 the uncertainties of the latter, except as regards the perihelion dis- 

 tance. Both comets were very favorably situated for determination 

 of the perihelion distance, and on mature consideration I am con- 

 vinced, that, unless it can be shown that the comet had been exposed 

 to perturbations, by the earth or Jupiter, capable of producing a very 

 great change in the perihelion distance, all arguments, drawn from the 

 similarity of the elements, in favor of the identity of the two comets, 

 must fall to the ground. In both cases, the comets approached quite 

 near the earth, and were observed in the ecliptic ; but in 1748 the 

 comet crossed this plane so far inside the earth's orbit, and in 1849 so 

 far outside of the same, that all attempts to attribute the discrepancy of 

 the perihelion distances to errors of observation or computation, in 

 either case, must be fruitless. 



" It must, nevertheless, be acknowledged that the resemblance of 

 the two orbits is greater than exists between those of any other two 

 comets on record. In order, therefore, to discover whether any indi- 

 cation of periodicity were to be found in the orbit itself, application 

 was made to Mr. Bond, of the Cambridge Observatory, for three ob- 

 servations, as remote from one another as possible ; and from observa- 

 tions on April 11th, 19th, and 27th, I computed an orbit by Gauss's 

 method, without any hypothesis whatever as to the nature of the conic 

 section described. The resulting curve was no ellipse at all, but the 

 following hyperbola. 



9> 30° 11' 11" 



i 67 19 39 



n 266 16 36 



V^ 10 6 1 



