ASTRONOMY: C. G. ABBOT 
313 
THE SMITHSONIAN 1 SOLAR CONSTANT EXPEDITION TO 
CALAMA, CHILE 
By C. G. Abbot 
Smithsonian Astrophysical Observatory 
Communicated August 16, 1918 
As early as 1903 the observations of the Smithsonian Astrophysical Ob- 
servatory suggested the view that the solar radiation varies over a range of 
several per cent within intervals as short as a few days or weeks. We were 
measuring the radiation of the sun at the earth's surface. The measurements 
comprised determinations of the heating effect of tbe solar rays on the black- 
ened surface of the pyrheliometer, which measures all rays of the spectrum as 
found in 'white light/ and also the heating effects at all parts of the solar 
spectrum as detected by the bolometer, including ultra-violet, visible and 
infra-red rays. We measured at intervals on every clear day as the sun 
declined from high altitudes near the zenith to low altitudes near the horizon. 
Thus the rays measured rjassed through longer and longer paths, according 
to the obliquity with which they crossed the atmospheric layers, and con- 
sequently they grew weaker and weaker as the sun declined lower and lower. 
From the spectro-bolometric measurements, standardized to calories by aid 
of the simultaneous pyrheliometer measurements and reduced to zero atmo- 
spheric absorption by the method of Langley we thus determined the intensity 
of solar radiation as it would be outside the atmosphere at mean solar dis- 
tance. This has been called the solar constant of radiation. Its average 
value is about 1.93 calories per square centimeter per minute. 
As I have said the results of 1903 at Washington indicated variations of 
this so-called constant over a range of nearly 10%. Owing to the prevalence 
of clouds at Washington all too few observations were available. Neverthe- 
less when we compared such as we had with the anomalies of temperature 
of the North Temperate Zone as represented by meteorological observations 
at 89 stations, all regions showed a variation of temperature nearly parallel 
to, and of a proper magnitude to correspond with, the supposed variations of 
the sun. 
In 1905 we transferred the observing to Mount Wilson, California, and 
with the exception of 1907 we have observed the 'solar constant,' in that 
relatively favorable place, usually from June to October of each year. The 
results have confirmed the apparent variability of the sun. It is impossible 
to go outside the atmosphere to observe, and we feared that the apparent 
variability of the sun might have been really due to defects in our estimation 
of the losses in the atmosphere. To check our work as far as possible we 
observed in 1908, 1909, and 1910 from the summit of Mount Whitney (4420 
meters) the highest peak in the (older) United States. No error dependent 
