OBSERVATION OP SUDDEN PHENOMENA. 
49 
its own complete recording apparatus and be more portable 
than an ordinary opera-glass.' 
I have not found time to use this apparatus on the moon 
or occultations, but I have, what is possibly more to the pur¬ 
pose now, tried it on an artificial star, the instant of whose 
appearance and disappearance was independently recorded 
on a chronograph by an electrical contact. Different ob¬ 
servers, entirely unskilled and ignorant in the use of the in- 
.strument, were invited to look into it and to determine the 
quadrant and section in which the star appeared and disap¬ 
peared. 
I have momentarily mislaid my notes containing in full 
detail the results of four observers, but I can summarize 
them approximately in saying that after being simply told 
what to note; the average probable error—(that is, for any 
single observation)—was rather less than one-twentieth of a 
second. As far as I can judge from the limited number of 
instances, the younger the observer the better the observa¬ 
tion. The worst of the observers (the oldest), however, had 
a probable error considerably less than one-tenth of a second; 
the youngest, a probable error of something like one-fortieth 
of a second, which implies, as you will observe, that he not 
only readily noted the quadrant and the subdivisions of the 
quadrant, but, also as a rule, even the part of the subdivision 
in which the star was first seen. None of these observers 
had so much as one hour’s practice. 
The plan in question is easily adapted to meridian obser¬ 
vations, but for these we have numerous plans for correcting 
personal equation, and the writer may also direct attention 
to the fact of the existence of a distinct device (Am. Journal 
of Science, July, 1877) which practically eliminates the per¬ 
sonal error in the very act of a transit observation. It is 
more elaborate than the present one, which is so simple that 
it may be useful even in longitude work with the transit, 
though its proper field seems to be the observation of sudden 
events; but, to whatever purpose it is applied, I beg leave to 
5—Bull. Phil. Soc., Wash., Vol. 11. 
