REPORT OF THE SECRETARY. 81 



the spectro-bolometrir apparatus, in i)la('e of tlie canvas shelters thus far used, and 

 with these, the improved churning device i)roposed, and the tent to shelter the tube 

 already ordered, it is conlidently expected to have the horizontal telescope in very 

 satisfactory condition during the coining liscal year. 



3. Investigations. 



Sun's posslbJe variahilitii. —Notahle progress has been made witli the researches you 

 have initiated on the amount of solar radiation and its absorption in the solar envel- 

 ope and in our atmosphere. Within the last seventeen months three independent 

 kinds of evidence have been collected here, pointing toward the conclusion that the 

 radiation supplied by the sun may perhai)s fluctuate within intervals of a few months 

 through ranges of nearly or quite 10 per cent, and that these fluctuations of solar 

 radiation may cause changes of temperature of several degrees centigrade nearly 

 simultaneously over the great continental areas of the world. Further evidence 

 must, however, be awaited to verify this important conclusion. 



The tliree kinds of evidence referred to are as follows: First, on all favorable days 

 the "solar constant" of radiation outside our atmosphere has been determined here, 

 and changes of about 10 per cent in the values obtained have been found which can 

 not be attributed to known causes. Second, the solar image formed by the horizon- 

 tal telescope has been examined with the spectro-bolometer to determine the absorp- 

 tion of radiation within the solar envelope itself. If we grant for argument's sake 

 that the rate of solar radiation outside our atmosphere fluctuates rapidly from time 

 to time, then as you have observed, the cause of this fluctuation can not reasonably 

 be a variability of the temperature of so immense a body as the sun itself, but nmst 

 rather be in a change in the absorption of a more or less opaque envelope surrounding 

 the sun. Accordingly the two researches I have mentioned are intimately associated, 

 for if we find a considerable increase in the rate of solar radiation outside our atmos- 

 phere we ought to find a corresponding decrease in the absorbing power of the solar 

 envelope. 



Such is in fact one of the most notable results of the year's work. In August, 

 September, and October, 1903, the observations of the " solar constant" of radiation 

 indicated that the rate of radiation was about 10 per cent below that observed in 

 February, 1904. On the other hand measurements of the absorption of the solar 

 envelope indicated considerably less absorption in February, 1904, than in Septem- 

 ber, 1903. 



The third kind of evidence of change in solar radiation is based on a study of the 

 temperature of the North Temperate Zone, as indicated by the Internationaler Deka- 

 denberichte published by the Kaiserliche ]\Iarine Deutsche Seewarte, and received 

 at the Observatory through the kindness of the Librarian of the United States 

 Weather Bureau. This publication gives the mean temperature at 8 a. m. for each 

 ten days at each one of about one hundred stations distributed over the principal 

 land areas of the North Temperate Zone, and for about ninety of these stations there 

 is also given the normal temperatures for the same ten-day periods, representing the 

 mean of many years. From these data there have been computed here the temper- 

 ature departures from the normal since January 1, 1903, and these are compared 

 graphically in the accompanying chart, Plate Yll/' with the measures of the solar 

 constant made in 1903. It will be seen that shortly after the observed fall of solar 

 radiation in March, 1903, a general fall of temperature occurred, which would be a 

 natural result of such a change. It has been shown here, in accordance with the 

 known laws of radiation, that 10 per cent fall in the solar radiation could not pro- 

 duce more than 7°. 5 C. fall in terrestrial temperatures, and that several causes, notably 



«Shown also in your article on "A Possible Variation of the Solar Radiation," 

 Astrophysical Journal, June, 1904. 

 SM 1904 6 



