344 A. D. Bache on Determining Latitudes by Talcott’s Method. 
stars was absolutely identical, with a probable error of + 0'"11 
for the result by either observer. A further trial with zenith 
telescopes No. 10 and No. 2 at Mount Harris station, Maine, ed 
assistant Dean and sub-assistant Goodfellow, under my directio 
in which Mr. Dean used both instruments, ‘and M 1. Goodfellow 
No. 10, gives latitudes differing very sligh tly. Mr. Dean ob- 
tained using zenith telescope No. “10, 44° 39’ 54/49, and by fae 0. 2, 
5476, while Mr. Goodfellow using ‘No. 10 obtained 54’83. 
corresponding probable errors of the mean latitudes from ‘hana 
irs of stars were + 0/14, 0-18, and 0-13. Finally, 
at Moet t Desert station, Maine, a comparison was made by 8u 
assistants Edward Goodfellow and Stephen Harris, using zenith 
telescope No. 5, and observing the same pairs of stars. e 
latitudes obtained from observations on thirty pairs of stars dif 
fered but two — of a second, with probable errors of 
+ 0°13 and+0'"1 
~The investigation Ais thus clearly proved that there is no per- 
sonal equation of observer or instrument in the use of the zenith 
telescope by Talcott’s method, but that identical results can 
obtained by different observers, with the same or different instru- 
ments using the same pairs of stars for observing. The accuracy, 
of the final. result, depends therefore, upon the accuracy of the 
declinations of the pairs of stars, upon the use of the proper 
number of pairs, aid upon the care taken in determining ed 
values of the fitoreiaaaee and level divisions, and in using 
instrument. A probable error in observing of from + 0/25 to 
+ 080 may be expected for one observation, according to the 
size and quality of the pe ees and the ability of the observer. 
The probable error of the result by a single pair of stars as de- 
posing on the catalogue anon of the stars places, will range 
m + 060 to + 1-00, which will give from + 011 to + 
018 for the latitude as resulting from thirty pairs. The dis- 
crepancy in the results obtained at Sebattis may posst ossibly 
owing to the warping of the thin ledge of rock on which the 
instruments were mounted. The unusually large value of the 
rity error of observation seems to indicate some distur 
influe 
The following table contains a recapitulatio on of these results. 
The column gives the name of the station at which the ob- 
ete pha 
