INTRODUCTION 
XV 
the coordinate when reduced to 1910. From inspection of this information one 
may determine what stars are best suited for use as standard stars, when both 
elements of qualification for standard stars are considered. The precision of the 
position at a given epoch is compounded of the precision at the principal, or mean, 
epoch with that of the proper-motion. If we construct a standard catalogue that 
is designed to remain in efficient use until 1920 (say), then its predicted probable 
errors during its period of use up to 1920 are of even greater interest at the later 
than for the earlier dates. The data given in the present Catalogue provide for the 
computation of a predicted probable error of position for any required epoch; and 
whenever the resulting probable error is not much over 0V3, this prediction as to 
precision should be relatively a fair approximation to the truth. 
This information, however, may frequently be required in relation to stars which 
would not now be regarded as well enough determined to be employed as standard 
stars in zone or regional observations at fixed observatories, but which might be 
made into useful secondary standards through special differential observations 
based upon the principal standards, and contemporaneously with the zone-observa¬ 
tions themselves. This method has been advantageously employed at the Dudley 
Observatory (1896-1901, A. /., 499). The data of the present Catalogue would 
facilitate the selection of such standards in such a manner that the new secondaries 
might be chosen in view of their future as well as present usefulness. Not only is 
the selection of such stars herein facilitated, but the preparation of accurately pre¬ 
dicted positions for them is made comparatively easy. The results contained in the 
present Catalogue are supposed to be exhaustive of existing determinations to the 
extent indicated in the foregoing list of catalogues employed. Additional weight of 
observations, not contained in that list, can easily be incorporated with the catalogue- 
data in such a manner (as will be explained hereafter) that the final results will 
virtually represent a complete solution of all the observational equations that were 
formed in the present work, together with those from the additional material. 
Thus the work here done, if considered satisfactory, need not be duplicated except 
in special cases. 
As another illustration of the value which an approximate knowledge of the 
degree of precision reached may have, take the selection of stars for determination of 
latitude. In the Catalogue the probable errors in declination are given for 1910; 
and they can easily be computed for other dates. This is of very great advantage 
in facilitating choice of the best determined stars. Furthermore, considering the 
great precision of instrumental results in latitude-observation, the precision of 
the final result of a series of latitude-observations may depend more upon the 
precision of the star-places than upon that of the instrument. The present Cata¬ 
logue furnishes fairly reliable estimates of probable error for each declination, 
enabling this element to be taken into account in latitude-reductions in order to 
reach the most probable result. At the same time the results of later observations 
of such latitude-stars can be easily and accurately incorporated in a revision of the 
determinations recorded in the present work, as will be shown later on. It should 
be specially noted that for stars of the class represented by the weaker one-third of 
the present Catalogue, the probable error of the predicted position increases much 
