1921] on Cloudland Studies 431 



proportions of the absorbable constituents (water vapour, carbon 

 dioxide and ozone) in the atmosphere of the chamber were controlled, 

 and the resulting alterations of intensity at all parts of the analysed 

 beam determined by the bolometer. The measurements up to wave 

 lengths of over 50 \x were finally extended with extraordinary 

 precautions to the atmospheric absorption of the sun's radiation. 



Lord Rayleigh, in a Discourse given in the Royal Institution on 

 May 7, 1920, gave an account of the absorption of the short ultra- 

 violet solar radiation by the thin layer of ozone in the upper air. 

 The absorption of ozone for the long wave lengths has been measured 

 by the charcoal thermoscope, the same apparatus as was employed 

 for the zenith observations being used. A dilute solution of liquid 

 ozone in liquid oxygen was placed in the central vessel A above the 

 thermoscope cell, the ozone solution being obtained by bubbling a 

 slow stream of ozonised oxygen into A. The diaphragm drum H 

 was removed, and the liquid was kept clear of solid impurity by an 

 electrified quartz tube. The displacements obtained by exposure to 

 the black source at room temperature were determined before and 

 after the condensation of liquid ozone in the liquid oxygen. The 

 percentage of ozone was afterwards determined by decanting out a 

 small measured quantity of the ozone solution in A and evaporating 

 it through potassium iodide. The observations had to be made with 

 the least possible delay, as even the very dilute ozone solution 

 employed soon attacked the rubber membrane. A succession of 

 crackling explosions occurred, which resulted each time in a loss of 

 concentration of the ozone, and finally the membrane was punctured. 

 A rock-salt cell would no doubt have been preferable, but was not 

 available at the time. 



Similar measurements with more concentrated ozone, or better 

 still with pure liquid ozone, would be interesting. For more con- 

 centrated solutions a separate vacuum vessel would be desirable, the 

 radiation to be measured being reflected in and out again by double 

 silver mirrors, before being finally reflected into the thermoscope. 

 This arrangement was tried with ordinary clean liquid oxygen and 

 found to be quite workable. The results obtained with the dilute 

 solution of ozone as described showed that 5 cm. depth of an ozone- 

 oxygen liquid, containing only 0' 18 gramme of ozone in the total 

 88 grammes of liquid oxygen, reduced the manometer reading by 

 16 per cent, on exposure to the same black source as the oxygen at 

 l'.>'7° C, the values being 6*0 cm. in half a minute with pure 

 oxygen, and 5*25 to 5*35 cm. in the same time with the 1 in 4*9 

 dilution of ozone-oxygen liquid. By volume the proportion would 

 be 1 of ozone to 730 of oxygen, so that the actual thickness of pure 

 ozone liquid if separated would be only 0*07 mm. Fowle and 

 Aldrich concluded that the ozone in the upper air cuts off one-fifth 

 of the outgoing radiation of the earth which has been transmitted 

 through the lower layers of the atmosphere. To the quality of 



