Vol. 6, 1920 
ASTRONOMY: C. G. ABBOT 
675 
causes of variation in addition to the supposed solar changes affect the 
brightness of Saturn. In the observations recently sent to me Dr. Guth- 
nick has corrected for all of these influences except those associated with 
phase. These produce quite a large effect, so that, for instance, the com- 
parisons between Saturn and a Leonis show gradually progressive differ- 
ences ranging from 0.61 to 0.90 stellar magnitude. He states that for 
1920 observations the influence of phase is approximately eliminated by 
subtracting from the given magnitude differences the product of the phase 
angle in degrees by a coefficient, which before opposition is 0.025, but 
after opposition is 0.039. These correcting coefficients are, however, 
inapplicable for observations made within less than 1° of opposition. 
Saturn then becomes considerably brighter than would be expected. 
I have reduced the observations on Saturn to constant phase angle by 
applying the coefficients just mentioned. On three dates, namely, Feb. 
25, Feb. 28 and March 1, the phase angle is less than 0.5°, so that the cor- 
rections are not applicable for these dates. The observation of March 13 
is stated by Dr. Guthnick to be of small weight because of unfavorable 
sky conditions in Berlin. Those of April 23 and May 12 are not of very 
great weight because Saturn was too near the horizon. 
In a first comparison I employed solar constant observations made in 
Chile on even dates with the observations made in Berlin. The result 
was so unsatisfactory as in my judgment to discredit the first hypothesis 
of a solar variation, appearing simultaneously in all directions. I then 
turned to the second hypothesis, making allowance for differences of 
heliocentric longitude of the earth and Saturn, and also for differences of 
terrestrial longitude and times of observing between Chile and Berlin. 
I thus arrived at the results given in the following table. In addition to 
the six days mentioned above as unsuitable, Feb. 7 is also unsatis- 
factory, owing to very cloudy conditions in Chile. 
There remain nine days suitable for comparison. These nine days are 
on tlie whole favorable to the view that the sun and Saturn each varied 
in brightness by two per cent, and that their variations were synchronous 
and proportional. The largest discrepancy which occurs among these 
nine comparisons is for March 9.5/10 and it indicates an error of slightly 
more than 0.5 per cent, or 0.005 magnitude. This amount of error is 
surely to be expected occasionally. Indeed it is very remarkable how 
accurate the results appear to be both in Berlin and Chile. 
Extraordinary March of Solar Radiation igiQ-20. — For about six years 
the intensity of solar radiation has almost always exceeded the mean 
value, 1.933 calories per square centimeter per minute, which was found 
from the Washington, Mt. Wilson and Mt. Whitney observations of 
1902-12, as published in Volume III of the Annals of the Astrophysical 
Observatory. This condition of affairs was expected to attend the return 
of increased solar activity, otherwise evidenced by numerous sunspots. 
