1902.] on Problems of thf. Atmosphere. 227 



and a temperature of — 220° ; B to 67 kilometres and — 181° C. ; C to 

 76 kilometres and - 132° ; and D to 87 kilometres and - 67°. 



On any of these temperature gradient hypotheses it appears that 

 practically above 56 miles the atmosphere would be substantially 

 composed of hydrogen. If helium and neon had been included in 

 the calculations they would have been found concentrated at high 

 elevation between the regions occupied respectively by the hydrogen 

 and the nitrogen in the diagrams. If the temperature is taken as 

 constant, Diagram II. shows that at an elevation of some 62 miles 

 the composition of a sample of the air, if it could be secured, would 

 be 95-1 per cent, of hydrogen, 4*6 per cent, of nitrogen, and 0'3 

 per cent, of oxygen. 



The permanence of the composition of the air at the highest 

 altitudes, as deduced from the basis of the dynamical theory of gases, 

 has been discussed by Stoney, Bryan, and others. It would appear 

 that there is a consensus of opinion that the rate at which gases like 

 hydrogen and helium could escape from the earth's atmosphere would 

 be excessively slow. Considering that to compensate any such loss 

 the same gases are being supplied by actions taking place in the 

 crust of the earth, we may safely regard them as necessarily per- 

 manent constituents of the upper air. 



The temperature at the elevations we have been discussing would 

 not be sufficient to cause any liquefaction of the nitrogen and oxygen, 

 on account of the j^ressure being so low. If we assume the mean 

 temperature as about the boiling-point of oxygen, then a considerable 

 amount of the carbonic acid must solidify as a mist, if the air from a 

 lower level be cooled to this temperature ; and the same result might 

 take place with other gases of relatively small volatility which occur 

 in air. The temperature of the upper air must be above that on 

 the vapour pressure curve corresponding to the barometric pressure 

 at the locality, otherwise liquid condensation must take place. In 

 other words, the temperature must be above the dew-point of air at 

 that place. At very high elevations, on any reasonable assumption 

 of temperature distribution, we inevitably reach a temperature where 

 the air would condense, just as Fourier and Poisson supposed it would, 

 unless the temperature is arrested in some way from approaching 

 the zero. 



Both ultra-violet absorption and the prevalence of electric storms 

 may have something to do with the maintenance of a higher mean 

 temperature than we should anticipate, following the deductions 

 of our assumed formulas for temperature decrements. The whole 

 mass of the air above 40 miles is not more than y^^th part of the 

 total mass of the atmosphere, so that any rain or snow of liquid 

 or solid air, if it did occur, would necessarily be of a very tenuous 

 description. In any case, the dense gases tend to accumulate in the 

 lower strata, and the lighter ones to predominate at the higher alti- 

 tudes, always assuming a steady state of equilibrium has been reached. 



It must be observed, however, that a sample of air taken at an 



