CHAPTER XIV. 
GENERAL CONCLUSIONS. 
I. PRECISION OF THE PHOTOMETER MEASURES. 
The tables headed “Standard Magnitude Stars” and “ Mean Magnitudes of 
Comparison Stars” furnish data for some important conclusions in regard to the 
precision of the photometric work. The method of reduction of the measures 
with the 6-inch, as explained on page 16, gives magnitude values for each of the 
standard stars which must average the same as the catalogue values. If both 
were perfect, the separate residuals between the catalogue and measured magni¬ 
tudes would be zero. Except for systematic differences, the actual residuals 
represent the combined effect of the accidental errors of the two systems com¬ 
pared. Table 117 collects the mean values of these residuals for the three standard 
stars in each field for each night, expressed in hundredths of a magnitude and 
taken without regard to sign. The mean residual for each of the three or four 
nights is given separately from the Harvard and from the Potsdam catalogue 
values. Then follow two columns giving the means of all the nights under the 
headings H and P for the two systems. The next two columns give the number 
of nights’ observations and the mean residuals of my measures among them¬ 
selves. Then follow the Potsdam colors, W standing for white, G for yellow, 
GW for yellowish white, WG for whitish yellow, the plus and minus signs indicat¬ 
ing a greater or less degree of color. 
Before these results are discussed, it is necessary to find if there is a systematic 
difference, due to color, between my measures and the two catalogues. The 
standard stars and their residuals were arranged according to color, and means 
taken of five groups each containing at least four stars, 28 stars in all. It was 
found that I see white stars fainter and yellow stars brighter than the catalogues, 
the range between white and whitish yellow stars being 0.20 as compared with 
the Harvard and 0.07 as compared with the Potsdam values. Correcting the 
28 stars for the color effect, the mean residual was reduced from 0.12 to o.n for 
the Harvard system, but remained at 0.06 for the Potsdam system, an effect so 
slight that it may be disregarded in the discussion. 
The results of Table 117 can now be discussed. 
(1) Considering the agreement of my measures among themselves .—The average 
difference between a single night’s magnitude of a star and the mean of the three 
(or in one case four) nights, ranges from 0.02 to 0.07 when the means for the 
three stars in the field are considered. The residuals for the separate stars 
range from 0.01 to 0.11; the general average of the residuals is ±0.04, correspond¬ 
ing to a probable error in the mean of a little less than 0.03 magnitude. The 
extreme values of the probable error are ±0.01 and ±0.05 for the separate stars. 
187 
