July 22, 1920] 



'ATI! RE 



645 



We are here dealing with the constrained motions 

 of falling bodies corresponding to those sliding on 

 inclined planes all joined at their highest points. 

 Such sliding motions are subject to the acceleration 

 of gravity, and hence are slow in starting, but later 

 attain high velocities. Since the falling body is air 

 which is displacing warmer, and hence lighter, air- 

 layers, in the case of the glacier its motions are 

 further modified as a result of adiabatic changes, 

 and, since large quantities of moisture are involved, 

 by important transformations of sensible and latent 

 heat. The source of this moisture is believed to be 

 largely the ice-needles of the cirri. 



The tendency to produce centrifugal surface-air 

 circulation above the glacier (anticyclonic movement) 

 is promoted by quiet conditions of the atmosphere, 

 since the measure of contact cooling of the surface 

 layer of air over the ice is a direct function of time. 

 The halting of this circulation or the induction of 

 any reverse centripetal movement of the surface air 

 (cyclonic movement) is an inverse function of the 

 time, since it is a direct function of the distance the 

 air currents descend vertically during their outward 



Fig. 2. — Device used to produce anticyclonic circula 

 in air above a cold dome. 



movement. Each of these movements is, however, 

 modified by the transformations of sensible and latent 

 heats of fusion and evaporation of the water brought 

 in in the form of the ice-needles of the cirri. 



The beginning of the glacial blizzard, slow by 

 reason of the flattish surface of the ice dome and the 

 acceleration of gravity, is also retarded by the neces- 

 sity of fusing and vaporising the ice-needles high up 

 in the vortex of the forming anticyclone, which causes 

 abstraction of heat and local displacements of air ; 

 whereas heat is evolved near the end of the blizzard, 

 when fresh snow is precipitated near the glacier sur- 

 face. Both these transformations of sensible and 

 latent heat will operate so as to add their effect 

 rather than to counteract that due to cooling or to 

 adiabatic effect. They thus tend to cause blizzards to 

 develop gradually and to end suddenly. The halt — the 

 end of the stroke of the refrigerating glacial engine — 

 comes about as soon as the rapid descent of the air 

 carried out by the blizzard has, through its adiabatic 

 effect, quite overcome the surface cooling due largelv 

 to the earlier calm. The length of the blizzard, if it 



NO. 2647. "^'OL. 105] 



precipitates fresh snow, should therefore be adjusted 

 in a measure to the expanse of the glacier surface 

 over w^hich the currents of air must slide before 

 gaining the two miles of descent on the dome, in 

 addition to that which takes place in the "eye" of 

 the anticyclone. Wm. H. Hobbs, 



Ann Arbor, Michigan, U.S.A., June 17. 



The Diamagnetism of Hydrogen. 



The fact quoted by Dr. Oxley in his letter tcr 

 Nature of July 8, that the diamagnetism of hydrogen 

 becomes less as the temperature is raised, seems to 

 be in favour of a kinetic hypothesis of the diamag- 

 netism of that gas rather than against it. 



If a magnet starting from rest is made to oscillate 

 it remains paramagnetic until the oscillations on either 

 side of the position of rest become 130°, after which 

 it behaves as a diamagnetic body, the diamagnetism 

 increasing until rotations begin. But once in rotation 

 the diamagnetism diminishes as the rotational energy 

 increases; and when this energy is very great the 

 magnet is nearly indifferent to a" magnetic field, and 

 it appears to be non-magnetic. If it is allowable to 

 treat temperature as a measure of this energv, then 

 this result means that the diamagnetism "should 

 become less as the temperature is raised, and this is 

 what has been observed. 



Since the paramagnetism of a rotating magnet is 

 found only for oscillations of less than 130°, the kinetic 

 energy must be comjjaratively small, and in the case 

 of hydrogen a change from diamagnetism to para- 

 magnetism can be expected to take place onlv when 

 the temperature is very near to the absolute" zero. 



Apart from the kinetic hypothesis, the fact that 

 there is any change at all "of the diamagnetism of 

 hydrogen with temperature is opposed to the accepted 

 view which regards true diamagnetism as independent 

 of tem]>erature. J. R. Ashworth. 



July 14. 



Occurrence of Ozone in the Atmosphere. 



With reference to the lecture of Lord Rayleigh 

 published in Nature of July 8 on "The Blue Sky 

 and the Optical Properties of Air," the conflicting 

 results obtained by chemical methods in the estima- 

 tion of atmospheric ozone are recalled. I beg to 

 direct attention to my. paper on "The Occurrence of 

 Ozone in the Upper Atmosphere" (Proc. Roy. Soc, 

 1914, A, vol. xc, p. 204), in which it is shown that 

 a reagent of potassium iodide solution can be made 

 to provide a basis for the distinction of ozone and 

 oxides of nitrogen at high dilutions and enable the 

 approximate estimation of the former. By this 

 method it is shown that, in accordance with the con- 

 clusions of Lord Rayleigh, ozone is present in the 

 upper atmosphere, the amount present at an altitude 

 of 10,000 ft. being of the order of 5 x lo-* parts f>er 

 unit volume. Measurements made with sounding- 

 balloons up to altitudes of 20 km. also showed the 

 presence of definite amounts of ozone, but no detect- 

 able increase between 4 km. and 20 km. The view 

 was put forward that this amount of ozone must be 

 taken into account in considering the optical pro- 

 j>erties of the sky. 



An extension of these measurements was made 

 with greater precision at the Mosso Laboratory on 

 Monte Rosa at an altitude of 15,000 ft., where an 

 average proportion of about ixio-* parts per volume 

 of ozone was found. J. N. Pring. 



The Victoria University of Manchester, 

 July 14. 



