REPORT OP THE SECRETARY. 89 



The somewhat paradoxical conclusions above stated seem capable 

 of explanation as follows: Associated with the great increase of 

 solar activity attending the maximum of the sun spot cycle, increased 

 convection is continually bringing fresh hot material to the sun's 

 surface, so that the effective solar temperature is then higher, and 

 greater emission of radiation prevails. At such a time the contrast, 

 which would be zero if the solar temperature were zero, is naturally 

 also increased. As for the quick, irregular fluctuations, it must be 

 supposed that the sun's outer envelope hinders somewhat the passage 

 of radiation from within outward. This hindrance is greater at the 

 edges of the sun's disk, where the path of the rays in the line of 

 sight is oblique, than it is at the center of the sun's disk. Suppose 

 now that the obstructive property of these layers varies from day to 

 day. When their transparency is increased the solar radiation must 

 increase; but as the effect will be most conspicuous at the edge of the 

 solar disk, where the path of the rays is longest, the contrast of 

 brightness between center and limb must thereby decrease. 



Two kinds of causes may, therefore, contribute to the sun's vari- 

 abilit}^ The one, a change of effective temperature attending the 

 general march of solar activity, may cause the variability of long 

 period. The other, a change of opacity of the outer solar layers, 

 may cause the variability of short irregular period. 



SUMMARY. 



Successful records of the intensity of solar radiation up to 25,000 

 meters were secured by means of automatic recording pyrhelio- 

 meters attached to sounding balloons. The mean of the three highest 

 values reduced to mean solar distance is 1.84 calories per square 

 centimeter per minute. Making 2 per cent allowance for scattering 

 and absorption in the air above (which gave a barometric pressure 

 only about one twenty-fifth of that at sea level), the value 1.88 

 calories is obtained as the probable intensity of solar radiation out- 

 side the atmosphere at mean solar distance on this day. This value 

 falls near the mean of numerous values obtained by spectrobolo- 

 metric observations on Mount Wilson. 



Experiments begun at sunrise and continued until 10 o'clock on 

 September 20 and 21, 1914, indicate great constancy of transparency 

 of the atmosphere above Mount Wilson, and yield solar-constant 

 values independent of the altitude of the sun. These results con- 

 firm the substantial accuracy of the Mount Wilson observations of 

 the solar constant of radiation. 



The radiation of the sun was 2.5 per cent below the mean, accord- 

 ing to the average of observations extending from July to October, 



