The Laws Observatory. 509 
written for each star, normal equations formed, and the most 
probable value of 47 and a deduced in the usual way. 
The average difference between the values of 47 deduced 
by these two methods is in this case less than 0°.005. 
The probable error of one 47 from a time star at Columbia 
as given by these observations is --0°.059 and at St. Louis it 
is +0°.053. The probable error of the clock correction as 
determined before or after exchange is for Columbia 05.015 
and for St. Louis +0°.019. Aside from the uncertainty due 
to personal equation, the probable error of the mean 4A from 
the five nights is about one-hundredth of a second. 
Personal Equation Observations. Observations for differ- 
ence of personal equation were made in St. Louis on Novem- 
ber 6,17 and 18, 1891. The Columbia transit instrument 
was taken to St. Louis and mounted on the eastern pier in 
the transit room of the Observatory of Washington Univer- 
sity, about five feet east of the pier on which the St. Louis 
transit instrument stands. The same stars were observed 
simultaneously by myself and Professor Pritchett, using the 
same clock but different chronographs. Observations of both 
time stars and polars were made in both positions of the instru- 
ments, and every precaution was taken in both the observa- 
tions and reductions for the accurate determination of the 
clock correction and the instrumental constants. The results 
given below are so discordant that they must be considered as 
giving only a rough approximation to the correct result. 
Clock AT. a; 
DATE 1891. P.—U.  |stars. 
Pritchett. Updegraff. 
November 6 +588.507 +58*.617 —0°.110 is 
Ky 17 +55 .836 | +55 .505 +0 .331 13 
es 18 +55 .870 | +55 .627 | +0 .243 11 
ee 18 +55 .806 | +55 sf18 +0 .093 9 
+0°.139 
Correction for dist. betwelen piers —0 .005 
+0°.134 
