14 
HAGEN. 
would like to mention that these discoveries and hypotheses 
seem to bring several branches of astronomy, apparently 
so different from each other, into close relation. The 
astronomy of binary stars would seem to comprise three 
principal classes: First, those systems whose components 
are far enough apart to be discerned visually; then, those 
whose components betray themselves in the variation of 
brightness or, photometrically , whether the variation be pro¬ 
duced by eclipses or by tidal fluctuations in the stellar 
atmospheres; lastly, the spectroscopic binaries, or those whose 
orbital motions are recognized by Doppler’s principle. 
3. Photometric Catalogues. —The latest photometric cata¬ 
logue published is the second part of the Potsdam Durch- 
musterung (PD.), which gives instructive comparisons with 
previous catalogues, viz., the Bonn Durchmusterung (BD.), 
the Oxford Uranometry (UO.), the Harvard Photometry 
(HP.), and the Harvard Photometric Revision (PR.). A 
few words on the Magnitudes and Colors may be interesting. 
It is well known that the northern BD. gives the magni¬ 
tudes of stars down to 9 M .5 with a rather indefinite limit, 
while the UO. and the HP. go as far as the 6th magnitude. 
The lower limit of the PR. is 9 M .0, and of the PD. 7 M .5, ac¬ 
cording to the BD. scale. The I. Part of the Potsdam 
catalogue comprises the stars from the equator to declina¬ 
tion + 20°, the II. from this parallel to declination -f 40°. 
We make only a passing remark on the internal agreement 
in the magnitudes of the PD. The two observers, Muller 
and Ivempf, have found that their systematic difference in 
estimating the brightness of the stars in the photometer has 
changed its sign: 
I. Part, M — K= + 0 M .02, 
II. “ “ = - 0 M .02. 
The quantity 0 M .02 is, however, so small that it is quite 
admissible to take the mean of their observations through¬ 
out. From their own residuals they found a probable error 
of ± 0 M ,04 for the catalogue magnitudes. 
