In this paper we extend the development of the theory of star catalog
examination, coordinate mean error evaluation, and proper motion mean
error assessment introduced in the first paper of this series. Specifi
cally, we have now shown how to compare positions of a star in differe
nt catalogs rather than just coordinates. Thus, the enlarged theory is
intrinsically two-dimensional. Further extensions are conceptually st
raightforward and should have many applications to comparable catalog-
to-catalog homogenization and systematic error elimination in multicol
or photometric work. In two dimensions, the special case of equal coor
dinate errors leads to the Rayleigh distribution for the prediction of
the angular distance apart distribution. This prediction is tested us
ing all the catalogs examined in Paper III and using the Southern Refe
rence Star (SRS) program source catalogs against a recompilation of th
e SRS completed in 1993. The conclusions regarding the relative superi
ority of the large-scale star catalogs, which were the focus of Paper
III, are borne out by the new testing. There is, however, a confusing
result: the southern hemisphere tests are always superior to the north
ern ones. Also, the anomalous AGK3 findings, for a catalog without sta
tistical adjustment in any meaningful sense, is explained.