Presidential Address. 
425 
I come now to the observations made during the year 1911, where the 
Eepsold method, accompanied by the use of the mechanically driven web, 
has been employed. Seven observers took part in the observations, which 
were primarily concentrated on a series of close circumpolar stars, but 
included also exact time determinations. Whenever possible the clock 
error was independently observed by two different observers on each 
observing night, so as to afford a direct comparison between the results 
obtained by the different observers. A discussion of these comparisons 
has yielded the following determinations of the personal discordances : — 
C. = +0-008 
A. P. = +0-007 
B. C. = -0-011 
A.W. = +0006 
M. = -0-005 
J.J. - -0008 
S. = +0-001 
The extreme discordance here shown is under two-hundredths of a 
second, while it is very doubtful whether this may not be a result of the 
inevitable accidental errors of the observations, which would be still 
further reduced had it been possible to obtain a larger number of 
comparisons. 
I have confined my attention so far to the aspect of personal equation 
as it affects rapidly moving stars, such as are always used for exact time 
determinations. The Eepsold method of observing was primarily devised 
for the latter purpose, a ad at the time it was brought into use at the Cape 
I was unacquainted with any attempts to extend its use to the slow-moving 
stars surrounding the pole. The accurate observations of these stars, 
especially in the direction of Right Ascension, presents special difficulties, 
while it is of special importance in that the poles form the principal 
reference points in the sky to which all exact observations are referred. 
Considerable doubt as to the positions of the principal southern circum- 
polars, as given in the almanacs, existed in my mind, and it was thought 
that the new Transit Circle, with its elaborate system of meridian marks, 
the high stability of which had been fully established, would soon serve to 
settle these doubts either one way or the other. 
A series of experimental determinations of azimuth was attempted 
during the winter months of 1907. In order to eliminate errors due to 
the tabular positions of the stars observed, the device was adopted of 
observing the same star at upper transit in the evening, and again twelve 
hours afterwards at lower transit in the morning, or vice versa. It soon 
