98 On Shooting Stars. 
noted in less than two hours, and Prof. Brandes remarks that they 
were obliged to leave many unrecorded. On most evenings howev- 
er, the number was much less. In comparing the different observa- 
tions, the first question which arises is, how shall we identify a me- 
teor; i. e. how can we know that the same meteor was seen by two 
individuals. In the first place, after allowing for difference of longi- 
tude, the times should evidently agree ; and it is unnecessary to com- 
pare any observations, where the times are not nearly coincident. 
Secondly, the lines of direction in which the observers saw the me- 
teor must intersect. To determine the latter question, Prof. Bran- 
des adopts the following simple expedient. Any three points are 
situated in the same plane, and consequently, a meteor must lie in 
the plane of a great circle, passing through the two places of obser- 
vation. Rectify the celestial globe for the time and place of obser- 
vation ; mark upon its horizon the direction in which the two places 
are situated with respect to each other; mark also the two points to 
which the two observers referred the disappearance of the meteor. 
These two points must lie in the same great circle with the two for- 
mer points on the horizon. Hold fast upon the two points on the 
ii a string passing through one of the apparent places on the 
globe, it should also pass through the other; and if the second ap- 
parent place varies considerably from the-string, the two observations 
cannot have been made upon the same meteor, or the observations 
are so inaccurate as to be useless. ‘This method with a large globe 
tan error of half a degree in assigning the place of a me- 
teor; Sand greater accuracy that this cannot be expected in observa- 
tions which are necessarily made without the aid of instruments. 
When the times of observation, and the lines of direction agree, 
it is highly probable that the two observers saw the same meteor. 
Other circumstances may, however, increase this probability. The 
size of the meteor should always be noted, whether a star of the 
first, second, &c. magnitude; any peculiarity of color, train, &c. 
When all these circumstances are accurately noted, they furnish the 
means of deciding upon coincident observations, almost to a cer- 
tainty. 
In performing the subsequent calculations, Prof. Brandes employ- 
ed the formule of Olbers. These formule are obtained in the fol- 
lowing manner: 
