NO. 5 SOLAR VARIATION AND FORECASTING ABBOT 9 



of course temperature varies if the sun does, but that the barometer 

 can only follow temperature, and must lag behind. Please to reflect 

 that, depending on locality, from lo to 50 per cent of solar radiation 

 outside the atmosphere is absorbed in the atmosphere by smoke, 

 haze, water vapor, and clouds. Since the atmosphere has a very 

 small capacity for heat, the heating effect of this tremendous energy 

 absorption is very quick in the atmosphere, compared to what it 

 would be in the ocean or 011 the solid earth. Sarmiento is about 

 100 miles from the Atlantic Ocean, in a very dry, clear region. 

 Rainfall increases in every direction from it except the north. Sup- 

 pose solar radiation increases. The surrounding air and the air at 

 Sarmiento both immediately grow warmer and expand, but the eft'ect 

 is much greater over the cloudy regions than at clear Sarmiento. 

 Hence, air flows from all around to Sarmiento and raises the ba- 

 rometer there. Similar action centers exist all over the world. They 

 vary in position with annual change of cloudiness and from other 

 causes, which Mr. Clayton's paper discusses. 



I now pass to consider certain other criticisms before taking up 

 reasons why we are convinced that the sun varies, and that our 

 observations give substantially the true picture. 



2. It has been pointed out by the late Professor Newcomb and 

 by Professor Marvin that there is no evidence of a permanent hot 

 or cold side of the sun. This accords with our results. Such a 

 condition sometimes exists for a few revolutions of the sun, but not 

 permanently. Hence, the solar rotation can be used only with greatest 

 circumspection as a period to forecast by. 



3. Professor Marvin has suggested that our recent observations 

 are badly prejudiced by a terrestrial i2-'month periodicity. I will 

 not say that there was absolutely nothing of the kind in Mount 

 Wilson observations, but I regard it as nearly or quite nonexistent 

 in later work. He has mistaken a real 11 -month periodicity in recent 

 years for a 12-month periodicity. Mr. Clayton discovered the 11- 

 month periodicity over a year ago and reported it to me. Figure 5 

 shows maxima in January, 1920, and September, 1923, an advance 

 of 4 months in 4 periods. Figure 6, which Mr. Clayton prepared, 

 shows the matter still clearer, because the short period solar fluctua- 

 tions have been removed by a usual process of smoothing, and we 

 see clearly that the maxima and the minima succeed one another by 

 ii-month intervals. Additional minima occurred in April, 1924, and 

 March, 1925, so that the 11 -monthly depression has clearly shown, 

 excepting in May, 1923, ever since the year 1918. 



