Meteors of November, 1868. 403 
two places of observation and any point of the meteor’s actual 
path, as the end, for example, we may passa plane. This 
plane, produced, intersects the celestial sphere in a great cir- 
: A : ; . 
heavens. 
If then we draw a great circle through the two points last 
named, and through either end of the track as observed 
place, this circle should pass through the corresponding point 
of the other observed path. Any deviation from this circle is 
due to some inaccuracy of observation. It is necessary to as- 
sume that one track is to be lengthened, or the other shortened, 
or that one or the other is to be changed in direction, or i 
position, until the two observed paths satisfy these conditions. 
Again, three or more observed paths if correct must, if pro- 
duced backward, pass through one and the same point on the 
globe, to wit, that point from which the meteor was moving. 
This point must, moreover, be above the horizon. Again, if 
i e 
well 
meteor. : 
_ Lam well aware that this method of treating the observa- 
tions throws special and undesirable responsibility upon the 
Computer. But in any method of ahh commonly prac- 
Used, good observations have large probable errors, amount- 
Ing often to several degrees. The computer must assume 
duty of judging the observations, and be held responsible for 
its faithtul discharge. 
The computer is, moreover, now in a condition to use to ad- 
Vantage any special points in the observations which he knows 
to be well determined, as the passage of the meteor near, or 
Over, a particular star. Or the peculiar position of one of th 
observers may far more than counterbalance any in in 
the location of the path. 
_A chart, constructed for the purpose by means of the tables 
of the Coast Survey, upon which are plotted the locations of the 
_ Observers, is in measuring distances and directions when- 
ver it gives sufficient accuracy. The figure on plate II. is 
_ ® reduction of this chart. 
