MOTION OF THE SOLAR SYSTEM. 
97 
In the first place it is to be remarked, that the result has been deduced from all the 
stars whose annual proper motions were found to be not less than 0"T, without any 
selection or rejection on account of obvious discrepancies. But it is manifest that if 
there is observed a general tendency to motion in a determinate direction, while in one 
or two instances the motion is in a nearly opposite direction, the presumption will 
be that in such exceptional cases the disagreement arises from the circumstance that 
the parallactic motion is masked and concealed by the relatively greater proper mo- 
tion of the star. In the second place, a few of the stars under consideration are very 
unfavourably situated for correct determination of the right ascension according to 
the method practised by Lacaille, and also for correct comparison by reason of the 
uncertainty of the computed precession; and in such cases a disagreement with the 
general result naturally gives rise to a suspicion of error in the determination. Now 
there are two stars, (3 and y 1 Octantis (Nos. 15 and 18 in the table) which are in 
those circumstances. The difference between the observed and hypothetical direc- 
tions of their motion, or is 158° 58'-0 in the case of the first, and 179° 13'-0 
in that of the second, showing in the latter case an apparent motion almost directly 
opposite to the parallactic motion due to the hypothesis. Both stars are also situated 
within 8° of the pole, so that the determination of their right ascensions by Lacaille’s 
method of equal altitudes must be liable to considerable uncertainty. For these 
reasons it may be concluded that the probable accuracy of the result will be increased 
by rejecting those two stars from the calculation. 
Omitting, therefore, the two equations 15 and 18, the sums of the squares and pro- 
ducts of the numerical quantities in the remaining seventy-nine are as follows : 
(nn)= 137120*9, («a) = 37T490, (bb) = 267010 
(ab) = — 5-9259, (an) = + 1 10-849, (bn)=- 64-282; 
which give the following solution, 
dA=— 2° 41'-8±4° 44'*9, 
<tt)= + 1° 48'-5±5° 36'-0 
2(T)= 28°25'-2. 
Here the corrections, both in right ascension and declination, are considerably dimi- 
nished, and it will be remarked that the former has changed sign and become sub- 
tractive instead of additive. The probable errors are also considerably reduced, and 
both corrections are now within the limits of the probable errors of Argelander’s 
determination. 
On applying the above corrections to the assumed values of A and D, we obtain, 
as the result of the calculation from seventy-nine stars, 
A = 257° 4 , *4+4° 44'-9, D=+34° 18'T±5° 36'’0. 
Another omission will diminish the corrections and probable errors still further. 
One of the stars in Henderson’s catalogue Ophiuchi, No. 80 in the subjoined table) 
MDCCCXLVIi. O 
