949 
we may conclude, that the mixing by convection currents is imper- 
fect; this increase however is rather irregular and smaller than 
ought to be expected without mixing. On account of the tncertainty 
of the determinations and the unknown percentage cf hydrogen it 
is not possible to calculate the degree of mixing. 
Still another consideration leads us to reckon for the higher strata 
with the possibility of a much smaller percentage of hydrogen. For 
a long time it has been doubted, on account of the kinetic theory 
of gases, that it would be possible for hydrogen to remain in the 
earth’s atmosphere ; since its presence in the lower strata has been 
put beyond doubt, the possibility has been urged that hydrogen is 
being produced continually at the surface, e.g. from mineral waters, 
ete. If this be right, it might be that in higher strata in consequence 
of the powerful ultra-violet radiation a combination with oxygen took 
place, and hence the percentage of hydrogen happened to be much 
lower than we should have to assume according to the gas-laws. 
In order to determine further the influence of the decrease in 
the percentage of hydrogen, we first calculated how the course of 
the rays would be if there were no hydrogen at all. The gaseous con- 
stituent, which then comes under consideration first, on account of 
its low molecular weight, is Helium. Starting from the figure 0.00015°/, 
Helium at tbe surface, as assumed Hann, we arrive at the following 
composition of the atmosphere : 
0 20 40 60 80 100 120 140 KM. 
N 78:03 SIS OMe SIE See) 2.2 
Ons 20:99 Lise 42e) 8.4 5.5 3.0 0.6 — 
Ar 0.94 0.5 0.2 
He = 0.00015 — — O.1 LG NSIS 75.6" IS 
As the change from the “nitrogen-atmosphere” to the “helium- 
atmosphere” occurs at much greater height than that to the ““hydrogen- 
atmosphere”, and moreover the velocity of sound varies more gradually, 
the sound rays now become higher and less curved. The results are: 
Pi: a H. D. 
42°18’ 120 218 
35° 8’ 125 193 
31°30’ 130 196 
The limit lies at about 190 KM. for rays with an initial angle of 34°. 
Hence, in order to arrive at a limit of 160 KM. we have to 
assume a small percentage of hydrogen in the higher strata. That 
this percentage ought to be very small appears from the calculation 
which, by way of test, was made with a percentage, 6 times 
