REPORT OF THE SECRETARY. 87 



critics have objected against the Mount Wilson measurements that 

 a progressive decrease of transparency occurs during the morning 

 hours, and especially during the period ordinarily used in our ob- 

 servations, so that our estimates of atmospheric transmission are 

 in their view too high, and our solar-constant values too low in con- 

 sequence. It has been suggested by one critic that the period during 

 which the zenith distance of the sun changes from 85° to 75° would 

 be more suitable for the work. 



To test this matter, observations were begun at sunrise on Septem- 

 ber 20 and 21, 1914, and continued until 10 o'clock, the usual clos- 

 ing time. These days were exceptionally clear and very dry, and 

 seemed well suited to give excellent solar-constant values. The 

 conditions of experiment, discussion of observations, and results are 

 given in full in the paper by Abbot, Fowle, and Aldrich above cited. 

 The principal results are these : No considerable difference in trans- 

 mission coefficients appeared whether these were based on the whole 

 morning's observations, on the range of air masses usually employed, 

 or on the range recommended by the critic above mentioned. Six 

 solar-constant values were derived for the two days, based on these 

 three different treatments of the data. All six values fall between 

 1.90 and 1.95 calories per square centimeter per minute, in good 

 agreement with values obtained as usual on other days. The ex- 

 periments confirm the view that the atmospheric transparency above 

 Mount Wilson is sufficiently uniform for the purposes of solar- 

 constant investigations. 



LONG-PERIOD VARIATION OF THE SUN. 



In the year 1913 the solar activity, as judged by the prevalence of 

 sun spots, was less than at any time for about a century. The mean 

 of all solar-constant values obtained at Mount Wilson from July to 

 October, 1913, inclusive, was 1.885 calories per square centimeter per 

 minute. This value falls 2.5 per cent below the mean value for the 

 years 1905 to 1912, which was 1.933 calories. 



Beginning September 9, 1913, observations of the distribution of 

 radiation along the diameter of the solar disk were secured on about 

 45 days of September, October, and November. These showed that 

 the increase (or contrast) of brightness of the center of the sun's 

 disk over that which prevails near the edge was less than that which 

 was found from Washington observations of the years 1905 to 1907. 



In the year 1914 the solar activity became distinctly greater than 

 in 1913. The number of spots, to be sure, was not great, but other 

 phenomena joined in showing that the period of maximum sun spots 

 was about to come. The mean of all solar-constant values obtained 

 at Mount Wilson from Jime to October, inclusive, was 1.950 calories. 



