SOLAR VARIATION AND WEATHER—ABBOT 123 
is cast into an intense spectrum, which comes to focus on the bo- 
lometer. The bolometer, originally invented about 1880 by Dr. Samuel 
P. Langley, is an electrical thermometer so sensitive that a change of 
a millionth of a degree in temperature can be registered. A clockwork 
causes the solar spectrum to drift slowly across the fine hairlike re- 
ceiver of the bolometer, and at the same time causes a photographic 
plate to drop slowly past the tiny spot of light reflected from the 
mirror of the magnetic-needle system of the sensitive galvanometer 
connected to the bolometer. Thus is produced in less than 10 minutes 
a bolograph, or curve showing the distribution of energy of radiation in 
the spectrum of the sun from far up in the ultraviolet to far down 
in the infrared. Several such energy curves are taken with appro- 
priate intervals during a morning as the sun rises higher and higher. 
A group of them is shown in figure 1. Simultaneously with each bolo- 
graph the total heating effect of the rays is measured outside the 
tunnel with an instrument called the pyrheliometer (heat-of-the-sun- 
meter). Also the altitude of the sun above the horizon is taken simul- 
taneously with the theodolite to indicate the slant thickness of the 
atmosphere. From this combination of observations it is possible 
to compute the intensity of the solar radiation as it is outside our 
atmosphere in free space at mean solar distance. This is the solar 
constant of radiation. 
DAILY VARIATIONS OF THE SOLAR CONSTANT 
For 25 years the Smithsonian Institution has been collecting daily 
measurements of the solar constant, when practicable, with a view to 
determining the march of the variations of the sun’s output of radia- 
tion. These fluctuations are small in percentage, rarely exceeding 1 
percent. Figure 2 gives the still smaller variations of the monthly 
mean solar-constant values, 1920-1939. It therefore requires very 
great accuracy of observing to disclose and evaluate them, hampered 
as we are by the superincumbent highly variable atmosphere. We are 
at a disadvantage compared to astronomers who measure variable 
stars, for they can compare the star investigated with other similar 
stars nearby, all of which suffer equal percentage losses of light from 
atmospheric hindrances. The sun is unique and can be compared 
with nothing near it in the sky. One can only compare an absolute 
solar measurement of today against an absolute solar measurement of 
tomorrow, trusting altogether to the accurate determination of atmos- 
pheric transmission on each day to make the measurements comparable. 
The Institution maintains three solar-constant observatories, two 
in the Northern and one in the Southern Hemisphere, all on high moun- 
tains in desert lands. The following table and summary shows how 
well the solar-constant daily measurements at great distances apart, 
