JANUAEY 3, 1913] 



SCIENCE 



37 



1890, which may be used as a test by which less 

 perfect measurements may be judged. By theory, 

 we have the criterion that 1/ni must always be a 

 fraction. Eejeeting all observations which do not 

 conform to this criterion, also midday observa- 

 tions which are affected by the usual midday 

 depression, the author compares the mean of the 

 two series of observations by Kimball with 

 Savglief's measures. The mean values are: 



e = 7, Savglief l/m = 0.647, 

 Kimball l/m = 0.786; 

 ^ = (1 + 7)''-«' X 0.94 = 3.606 (Sav<ilief ) ; 

 A={1 + 7)"" X 0.72 = 8.698 (Kimball). 



Application of the criterion brings these other- 

 wise discordant observations into agreement. 

 Sky Radiation and the Isothermal Layer: F. W. 



Very. 



The author describes the mechanism of heat 

 transference and radiation absorption in the at- 

 mosphere, referring to his work on ' ' Atmospheric 

 Radiation," pp. 114-115 and 124. Vapors are 

 found in the atmosphere just so long as there 

 remains a quantity of even the least absorbing 

 gases sufficient to preserve the temperature above 

 the vaporization point. It was pointed out in dis- 

 cussing the absorbing power of gases and vapors 

 that the absorbing power of some gases was so 

 feeble that many miles of absorbing layers were 

 necessary to show the si^ectrum lines, while other 

 lines, notably those of water vapor, are produced 

 by the substances in very attenuated form. Water 

 vapor is found at the altitude of 30 km., and since 

 temperature inversion in the atmosphere is found 

 everywhere associated with excessive amounts of 

 aqueous vapor, there is no reason to doubt that 

 the great region of permanent temperature in- 

 version, the "isothermal layer," is due to this 

 substance. The extreme tenuity of the vapor 

 offers no difficulty to this supposition, since the 

 layer is many kilometers deep. I have found 

 the great Xi bands of aqueous vapor producing 

 almost complete extinction of the solar spectrum 

 through a range of 3 microns, when the tempera- 

 ture was — 30° C. This temperature is sometimes 

 exceeded even in the isothermal layer, where also 

 a relative humidity as high as 50 per cent, has 

 been observed. 



There are three principal loci of terrestrial 

 radiation, namely: (1) a thin, heated, superficial 

 layer of the terraqueous globe; (2) the solid or 

 liquid particles suspended in the air; and (3) the 

 upper static layer of air, called the isothermal 



layer, in which the permanent temperature-inver- 

 sion resides. The author 's supposition that ele- 

 vated regions of the atmospheric, up to something 

 like 20 km., undergo much larger temperature- 

 variations than the more deeply situated layers 

 of the free air, has been confirmed by Blair's 

 observations. This phenomenon also is attributed 

 principally to absorption by aqueous vapors. 



The earth 's effective temperature is not that of 

 an elevated layer of cool air, shown on the records 

 of the meteorological stations, but a mean in 

 varying proportions of the three loci named, which 

 owe their temperature to a complex of radiation, 

 convection, absorption, cloud precipitation, etc. 



The general conclusion from sky measurements 

 is that the effective sky-temperature is lower, on 

 the whole, when the dew-point is lower, and the 

 aqueous obstruction of radiation to space is least; 

 but even in the coldest and driest weather, the 

 sky temperature has never been found lower than 

 that of the isothermal layer, which behaves like 

 a nearly black body of approximate temperature, 

 2*^218° Abs. C, X max = 13.5 /I, and is opaque 

 to most of the radiation from the solid earth of 

 greater wave-length than this maximum. 



Orbits of the Visual and Spectroscopic Binary 



Star Epsilmi Sydrw AB : E. G. Aitken. 



The close double star known as e Hydra AB, 

 discovered by Schiaparelli in 1888, is unique 

 among the visual binary systems in that the ele- 

 ments of its orbit can also be deduced from the 

 spectrographie measures of the radial velocity. 



A revised system of elements from the microm- 

 eter measures, which now cover an arc of 450°, 

 gives 15.3 years as the best value for the revolu- 

 tion period. Adopting this as correct, the remain- 

 ing elements were computed independently from 

 the radial velocity measures, which extend from 

 1899 to 1911 and include both maximum and mini- 

 mum values. It was possible to find a set of ele- 

 ments which would satisfy both series of observa- 

 tions within the error of observation. These ele- 

 ments are: P = 15.3 years = 5,588 days; T = 

 1900.97 = .I.D., 2,415,375; e=0.65; w = 90°.0; 

 i = — 49°.95; fi = 104°.4; o = 0".23; a sin i — 

 493,000,000 km.; ■« = + 36.78 km.; S: = 8.45. 



From these elements we find the mean distance 

 between the two components of e Hydras to be 

 1,359,000,000 km., or 9.1 astronomical units, the 

 parallax of the system, 0".025 and its mass, 3.33 

 times the sun's mass. Seeliger has shown that the 

 periodic variations in the micrometer measures 

 connecting the third star C with the close pair 



