Observations of the comet of 1807. 3 
from the mean equinox. The right ascensions and declinations of 
the stars, made use of in reducing the observations, were taken from 
Zach’s catalogue, published in Vince’s Astronomy, and from Mas- 
kelyne’s, published in his “ Requisite Tab a 
Early in November I began to calculate the elements of the orbit 
by the method given by La Place, in vol. i. book ii. § 37 of the “ Me- 
canique Celeste.” For this purpose I made use of all the observa- 
tions from the first appearance of the comet to the sixth of November, 
as they are given in the table at the end of this article. | To render 
this part of the calculation more simple, and to avoid divisions by frac- 
tions of a day, I reduced all the seaervateies to seven —_ in ag 
pina iby Rabe to the observed } | = © 
ether gee 
the time of — and seven o hid 
This correction was found sufficiently near for this purpose,’ by tak- 
ing a proportional part of the daily variation of the comet in longi- 
tude and latitude. The observations having been thus prepared, I 
combined. them together in ten different examples, each one contain- 
ing five observations. By the mean of these calculations I found, for 
a first approximation, that. the time of passing the perihelion was Sep- 
tember 20,.1807, and the perihelion distance 0°65636, the mean dis- 
tance of the sun from the earth being estimated as unity. 
__. These elements I afterward corrected by the method explained by 
: La Place in the same section of his work. For this pur ose I made 
choice of the observations of October 8, October 23, and November 
6. From these I found, for a second approximation, that the correct. 
ed time of passing the "perihelion was September 19d. 7h. mean time 
“at Salem, and the perihelion distance 0°6645. These elements, with 
the longitude of the node, the longitude « of the Apel gs and: srdeg in- 
clination of the orbit to the eciiptic corres , Were publis! 
‘the Salem Gazette of the tenth of November 1907: in on it was 
