Mathematical discussion of the effects of radiation in the atmosphere is complicated by the presence of air and water va pour in varying proportions. The dry air is of little consequence in producing absorption or radiation, but as it shares with the water vapour the heat which the latter absorbs, its presence is equivalent to weighting the water vapour with an added specific heat.
While the lapse-rate re mains on the average less than the dry adiabatic, but slightly greater than the saturated adiabatic, we can readily conceive of any isolated mass of air which has become saturated and at a slightly higher temperature than its environment, being able to rise through its environment, since in the circumstances postulated its temperature would be at each successive level higher than that of its immediate environment. The converse process of the descent of air, however, is not readily understandable. This prob lem has not yet been satisfactorily solved, probably because the time scale of radiation effects has not been clearly laid down, and until this problem is solved it will not be possible to evolve a satisfactory explanation of the general circulation of the atmos phere. The direct effect of solar radiation is an ascent of air in the doldrums. The air which ascends flows away from the equator and it is of vital importance to determine at what stage in its flow poleward such air will be in a suitable condition to descend again to the surface of the earth.
The existence of ascending currents on a large scale is shown by the occurrence of rainfall, and it is thus possible to obtain some idea of the extent and intensity of such currents. Descending cur rents are usually less rapid, and in consequence their physical effects may be largely masked by the horizontal motions of the air. The outstanding feature of the distribution of temperature in the troposphere is its approximation to horizontal stratification, which is not a condition favourable to the descent of large masses of air from high levels to the ground. Moreover it appears prob able that the conditions existing in the upper air do not show the same variability from place to place which characterises the surface layers.
The amount of energy which can be made available by vertical motion in particular cases can be evaluated by means of the tephigram (vide Shaw, Manual of Meteorology, Vols. i, iii). For
an important discussion of the use of the wet bulb temperature in connection with the thermodynamics of vertical motion reference should be made to a paper by Normand (Memoirs Indian Met. Dept., Vol. xxiii).
It has been shown by Jeffreys (Q..1 .R. Met. Soc., 1926) and Brunt (Phil. Mag., 1926) that if no external source of energy were effective the atmosphere would be brought to rest relative to the earth in about io days by friction. Brunt has shown that the energy destroyed by friction could be made good by the conversion of about 2 per cent of the energy of the incoming solar radiation into kinetic energy. This may in part be brought about by the ascent of heated air in the doldrums, followed by motion poleward and descent in latitude 35° North or South in consequence of cooling, but the physical processes involved in such a cycle are not clearly understood (see Dictionary of Applied Physics, vol. iii, article by N. Shaw). Further the general circulation of the atmosphere has been shown to be a very complex phenomenon, lacking symmetry about the earth's axis, so that no simple theory can be expected to explain it. It is now considered that the cy clones and anticyclones form an essential part of the general circulation, (see Jeffreys Q.J.R. Met. Soc., 1926, 1927, and Exner Dynamische Meteorologie, 2nd edition, chap. ix). The distribu tion of temperature in the upper atmosphere, in particular the fact that the stratosphere is highest and coldest over the equator, is bound up with the general problem of the circulation of the atmosphere. For a discussion of some aspects of the stratospheric distribution see various papers by Simpson in the Memoirs R. Met. Soc., 1928, 1929. For a discussion of the present position of theories of the general circulation of the atmosphere by E. W. Barlow see Quarterly Journal R. Met. Soc., Jan. 1931; see also a paper by T. Bergeron in Meteorologische Zeitschrift, 193o. The work of Dobson and others on the distribution of Ozone (see Pro ceedings of the Royal Society, 1929, 193o) suggests a new line of attack upon some problems associated with the general cir culation.