Septembee 18, 1914] 



SCIENCE 



421 



the insolation in tlie middle of tbe day, and 

 over water surfaces not at all. Taking the 

 earth as a whole, therefore, the adopted esti- 

 mate of mean surface temperature, as agree- 

 ing with the mean local air temperature, can 

 not be altered more than a fraction of a degree 

 by considering the high temperatures of 

 strongly insolated rock and arid soil, because 

 the area occupied by such surfaces is small 

 compared with the vast expanses of ocean, 

 moist soil, and soil protected by vegetation, 

 which are not thus affected. 



Mr. Angstrom thinks that, of the terrestrial 

 radiation which escapes absorption by the 

 lower layers of the atmosphere, " a consider- 

 able part will be absorbed by the ozone in the 

 higher and colder strata of the atmosphere," 

 and he assigns 20 per cent, of the total re- 

 maining radiation as a probable value of this 

 absorption. Now if we note that the ozone 

 band covers not one tenth of the entire spec- 

 trum, and that 20 per cent, would be a fair 

 value for ozone absorption within the limits 

 of the hand (at least in summer), we may 

 conclude that the absorption which it exerts 

 is nearer to 2 than to 20 per cent, of the 

 entire spectrum. An example will confirm 

 this approximate statement: 



On several occasions of strong ozone ab- 

 sorption, the energy in the solar spectrum of 

 wave-length greater than great 9 (and for 

 our present purpose it makes little difference 

 whether a curve of solar radiation, or one of 

 terrestrial radiation be taken in this part of 

 the spectrum) had a mean value of 537 arbi- 

 trary units, as measured on a plotting of the 

 spectral energy-curve. On the same scale, the 

 area covered by the ozone band was equal 

 to IY.4 units, or the ozone absorption was 

 3.24 per cent, of the spectrum lying beyond 

 the center of the greatest of the bands of 

 aqueous vapor. The following separate values 

 show the variability of the band on days of 

 strongest ozone absorption ; (a) = ozone ab- 

 sorption in the band from 9.1 /a to 10.0 /* as a 

 percentage of the entire unabsorbed energy 

 between 6 /* and 20 /j,, (&) = ozone absorption 

 of the original unabsorbed energy within the 

 approximate and apparent limits of the band: 



Ozone Absorption (a) (&) 



Per Cent. Per Cent. 



(3.00 38.8 



3.33 36.9 



3.72 50.0 



3.29 34.5 



3.50 33.3 



A single day in July 2.59 21.0 



Mean 3.24 35.8 



The ozone absorption is considerably smaller 

 in summer than in winter, and ozone probably 

 has its greatest efficiency as a preserver of the 

 earth's heat in the polar regions. 



The conditions in my measures of sky radia- 

 tion were such that the surface temperature 

 could not have differed much from the adopted 

 air temperature, because in winter the sun 

 was low, and in summer the ground was moist; 

 but possibly the values assigned for unob- 

 structed radiation should be lowered to allow 

 for the diminished value of the earth's radia- 

 tive quality below that of a perfect radiator. 

 This, however, would increase the transmis- 

 sion, since the instrument with which the sky 

 radiation was measured was a complete 

 radiator. The thermopile had its very small, 

 blackened, absorbent surface at the center of 

 a hemispherical mirror, 10 cm. in diameter, 

 gold-plated and burnished, the rays entering 

 through a 1 cm. circular, central aperture, 

 entirely open to the outside air. Any rays re- 

 fleeted from the front surface of the thermo- 

 pile and falling on the mirror, were returned 

 back repeatedly for absorption. In spite of 

 the protection afforded by the case (and by 

 still another, but a wider aperture 2 m. in 

 front of the measuring surface), it was diffi- 

 cult to keep the instrument balanced during 

 very cold or windy weather. Measures were 

 taken in series of five readings. Unless clouds 

 interfered, these readings commonly agreed 

 to the extent indicated in the following ex- 

 amples : 



(1) Feb. 3, 1909, 10i'20'° to 10^30" a.m. 

 External temperature, + 19.°0 F. Dew-point, 

 -1-13.°5 F. Eelative humidity 76.5 per cent. 

 Wind, fresh W.S.W. Sky, milky blue. A few 

 remnants of dissolving strato cumuli low in 

 the east. 



(2) Feb. 3, 1909, 9" 30™ to 9" 40" a.m. 

 Temperature of snow-covered ground (ther- 



