XXX PRELIMINARY GENERAL CATALOGUE OF STARS FOR 1900 
This excellent agreement between the results from right-ascension and declination 
tends still further to make it seem reasonable that the final result for An is not 
seriously affected by anomalous systematic motions of the stars. 
The absolute term of the second equation of (C), which contains A E a alone 
(correction to assumed motion of the equinox), is taken from Newcomb. At page 88 
of his summary, Astronomical Constants , Professor Newcomb gives +"30 as the 
most probable common correction of the centennial motions of the stars in his 
former determination, Ni. Later, in another way, he arrives at an identical conclu¬ 
sion ( The Processional Constant , p. 71). The details making up these values are, of 
course, very discordant. The probable error must be fully ± '.'30, and may be ma¬ 
terially greater. In order to incorporate a correction for the planetary precession 
in the discussion, we add the equation, AX = "00, which means that there has been 
no revision of Newcomb’s value of X published in Astronomical Constants. The 
careful determination of L. Struve {Best, der Const, der Praec., 1887) would give, as 
the approximate correction of Newcomb’s centennial X, +'.'21; but the opportunity 
of selecting improved values of the masses of the planets at the time of Newcomb’s 
computation seems to make it inadvisable to give weight to L. Struve’s determina¬ 
tion in combination with Newcomb’s. 
Solving the equations (C), first forming the normal equations, we have the 
values of the unknowns already cited (see also under “Comp,” equations (C)), 
as those which best harmonize with all the facts. The large correction (+”79) 
of Newcomb’s motion of the equinox, N b would not generally have been anticipated, 
perhaps. But from the observations of Greenwich extending from 1835 t0 1895, 
Newcomb found the “approximate value,” +'.'5 {The Precessional Constant , p. 71); 
and at the same time he strongly urges the value of this correction which he had 
deduced from the right-ascension of Mercury, + To; so that the above value, '.'79, 
may not be considered altogether unreasonable. The deduced value of X lies be¬ 
tween that of Newcomb and that of L. Struve, as if the weight 3 had been assigned 
to the former and weight 1 to the latter. The value of centennial luni-solar pre¬ 
cession is + "58 greater than that of Newcomb and '56 less than that of Struve 
(Peters’s) and "13 less than that in Struve’s original paper. 
These quantities, derived from the discussion now in progress at the Dudley 
Observatory, are regarded, however, as provisional; and are primarily designed to 
be employed in the further progress of the discussion mentioned, as giving sys¬ 
tematically more reliable values of the proper-motion than those contained in the 
Catalogue. At a later time, it is hoped that this discussion may be resumed with a 
far greater and systematically more accurate stock of proper-motions. 
PROBABLE ERRORS. 
The three columns (in one space) on the left-hand page under the caption “prob. 
errors,” and the corresponding columns on the right-hand page, under the same 
heading, contain certain probable errors to which allusion has already been made. 
The first of the three on the left-hand page under the special caption, aEp, gives 
in hundredths of a second of arc the probable error of the catalogue right-ascension 
reduced to the mean epoch of observation as printed in its appropriate column. The 
