REPORT OP THE SECRETARY. 91 



temperatures all over the world. The correlations were positive in 

 and near the Tropics, negative in temperate zones, and positive near 

 the poles. A lag of from 1 to 5 days occurred, the lag being less for 

 tropical zones. The barometric pressure also appeared to join in the 

 correlations. By an ingenius application of his method Dr. Clayton 

 shows that the short interval fluctuations of solar radiation are not 

 altogether without periodicity, for the changes tend to repeat them- 

 selves after 11 and 22 days, respectively. The same tendency is 

 found in the temperature records of Buenos Aires. We are now en- 

 gaged in testing this conclusion by computations for other years. 



Computations of Mount Wilson solar observations went on in the 

 hands of Miss Graves as usual at Washington, and the computing is 

 practically up to date. 



Mr. Fowle's research on the effect of water vapor and carbon 

 dioxide of the atmosphere to absorb long-wave raj^s, such as the earth 

 sends out, is now ready for publication. Many of the best observa- 

 tions were made by him during the past year. Some observations 

 made in February, 1917, at a time when the humidity of the atmos- 

 phere was very small, proved of special value. Opportunity was 

 taken of using some of the apparatus prepared for the South Amer- 

 ican expedition to aid in making holographic observations on the 

 solar spectrum at very great wave lengths, reaching to 17 microns. 

 By means of the spectro-bolometer prepared for South America it 

 was possible to determine accurately the quantities of water vapor 

 in the path of the solar beam. 



Certain conclusions stated in Volume II of the Annals of the Astro- 

 physical Observatory may now be corrected to correspond with the 

 new information. We stated: 



We can by no means admit that the radiation from the solid and liquid sur- 

 face of the earth passes unhindered to space. * * * The clouds, whose 

 average presence includes 52 per cent of the time, * * * are even more 

 efficient screens to the radiation of the earth than they are to the radiation of 

 the sun, so during 52 per cent of the time we may regard the radiation of the 

 solid and liquid earth to space as zero. During the remainder of the time 

 water vapor presents almost as effective a screen * * *. From the com- 

 bined work of Rubens and Aschkinass, Langley, Keeler and Very, and Nichols, 

 we * * * conclude that a tenth part of the average amount of water vapor 

 in the vertical column of atmosphere above sea level is enough to absorb more 

 than half of the radiation of the earth to space, and it is highly probable that, 

 considering the greater air mass attending the oblique passage of many of the 

 rays to space, nine-tenths of the radiation of the solid and liquid surface of 

 the earth is absorbed by the water vapor of the atmosphere even on clear days. 

 On cloudy days none is transmitted, so that the average escape of radiation 

 from the earth's surface to space probably does not exceed 5 per cent. 



Some writers have attributed a large share of the absorption of the atmos- 

 phere to the carbonic-acid gas which it contains, but * * * in atmospheric 

 conditions the absorption of carbonic acid gas in the spectrum of the earth 

 appears to be confined to two bands extending from wave lengths 3.6 to 5.4/*. 



