DETERMINATION OF CONSTANT OF NUTATION. 131 
Greenwich observations themselves. I fully admit its weight, 
but must remark, Ist, that for the star common to our compu- 
tations, the maximum of aberration obtained differs too widely, 
in one case more than a second; secondly, that the mean of 
20:5038 
Richardson and De Lambre gives, __225, almost identical with 
4 
20°370 
Brinkley and Struvé 20350; and, thirdly, that the use of Mr.R.’s 
20°360 
constant would scarcely have changed my-result. In the case of - 
y Draconis, the most important in my list, I performed the 
computation with this value, and the change it produced was 
only >,55 of a second. , 
As to the casual errors depending on the circumstances of 
observation, I find for this star, that the probable error of one 
observation = + 096, and therefore, by the theory of proba- 
bilities, it comes out an even het, that. as far as such errors are 
concerned the result given by it is not uncertain to 0°"04, and 
Lyre has nearly the same probable error. Therefore, the slight 
discordances in my results proceed from other, and, as yet, un- 
known causes. Similar and greater discrepancies occur in Mr. 
Richardgson’s investigations, but it is curious that in Brinkley’s 
researches on nutation, Capella, and « Cygni give results less 
than the mean; « Lyre, 6 Tauri, and Castor above it. 
It might seem that a more accurate conclusion is attainable 
by assuming the proper motion of the stars as known from 
comparison of Bradley’s observations with those of recent date. 
This supposition would give the constant 9°181, 0°05 less than 
that given above; but I think it inadmissible, for these motions 
may not be uniform, and there may be changes in the instru- 
ment, the refraction, the observer, nay, even in the direction of 
gravity, as affected by local circumstances, which are functions 
of the time. Something of this sort does actually appear here. 
It is well known that Pond latterly believed in the existence of 
a general southern motion of the stars ; and though Brinkley has 
shown most fully that this is imaginary, yet it is remarkable 
that the corrections of Bessel’s proper motions, which my work 
has given, are, except in one instance, all negative. I infer from 
this that the Greenwich circle is undergoing some progressive 
change of figure, which makes it show polar distances too great 
for about 30 degrees south of the zenith; but if this be the case 
it is not likely long to elude the sagacity of Mr. Airy. 
The declinations which I obtain from these Greenwich obser- 
vations differ considerably from those deduced by Pond himself, 
K 2 
