78 EEPOKT OF THE ACTING SECRETAET. 



January the numbers rose very high, and the results became as high or higher 

 than any before recorded and remained through most of February fully 15 

 per cent above the Mount Wilson values of late August. After this there was 

 again a falling off, and the Mount Wilson work of May and June, 1906, gives 

 about the same results as the corresponding months of 1905. By anticipation 

 it may be added that July, 1906, does not appear to run parallel to July, 1905, 

 by bringing higher values. 



On the whole the solar-radiation work of the past year furnishes the strong- 

 est evidence of solar variability yet recorded by this Observatoi'y. 



E. Confirmatory indications of solar variahility. 



While lack of knowledge of the conditions surrounding the sun makes con- 

 clusions drawn from a study of the apparent absarption of the solar envelope 

 doubtful, still the following inferences seem to be reasonable. If the solar 

 envelope should decrease in transmissibility, the solar radiation ought at first 

 to diminish, but after a time the radiating substance behind the absorbing 

 envelope should increase in temperature, owing to the obstruction of its radia- 

 tion by the absorbing envelope, so that at length the solar radiation should 

 reach nearly its former value despite the greater absorption in the envelope. 

 Similarly a falling off in salar absorption should immediately produce 

 increased radiation and afterwards a decrease of radiation. 



The observations of the solar image made at Washington and Mount Wilson 

 in the past fiscal year are not yet so completely reduced as to be profitably com- 

 pared with the observations of solar radiation on the basis of this hypothesis. 

 But it may be noted that during the month of July, while solar radiation 

 appeared to be increasing, the transparency of the solar envelope appeared to 

 be increasing also, which is in accordance with expectation. Furthermore, 

 the values representing the solar transmission at this time were below the 

 mean values for several years, which also accord with expectation. 



Lack of knowledge of the local conditions which govern the earth's tem- 

 perature prevents us from making at the present time accurate forecasts of 

 the effect of a variation of the sun on the earth's climate. Thus, for example, 

 a decrease of solar radiation may cause in some localities a decrease of cloudi- 

 ness sufficient to allow as much sunlight to reach the earth as came before the 

 fall of solar radiation took place. For the world in general, however, it might 

 naturally be expected that a decrease of solar radiation would cause a fall of 

 the earth's temperature. It is therefore in accordance with expectation that 

 the summer and early winter of 1905 were cooler than the average, while in 

 January and February, 1906, there was uncommon mildness, followed by a 

 cool spring. 



SUMMAEY. 



The work of the Astrophysical Observatory for the year has principally con- 

 sisted in the continuation at Washington and on Mount Wilson of researches 

 designed to discover any variability of solar radiation. The results of the year's 

 work have furnished the strongest evidence yet secured that the solar radiation 

 reaching the limits of the earth's atmosphere varies frequently and notably in 

 amount. According to present information, the mean value of the solar constant 

 of radiation is not far from 2.12 calories per square centimeter per minute ; its 

 range of fluctuation is irregular and sometimes reaches 15 per cent, and its 

 periods of fluctuation are variable. 



C. G. Abbot, 

 Acting Director Astrophysical Ohservaiory. 

 Mr. Richard Rathbun, 



Acting Secretary of the Smithsonian Institution 



