937 
in some cases at least, to the change in composition of the atmos- 
phere, which is caused by the unequal decrease of the partial 
pressures of the constituents of the atmosphere. 
If no mixing by convective currents oceurred, each of the gaseous 
constituents of the atmosphere would form, entirely according to 
its own laws, an atmosphere, the pressure p of which at the height 
h would be given approximately by a formula such as 
ule 
HT 
In consequence hereof at great height the denser gases can only 
occur to a very small percentage, and the lighter constituents, of 
which hydrogen is the most generally known, must gradually begin to 
log p = log p, — 
domineer. The convection-currents alter this state of affairs only so 
far as the lower atmosphere (the troposphere) is concerned. Above 
10 or 11 KM. (at least in the temperate zone) little convection 
occurs, and above this level the change in composition begins never- 
theless. Above that same level the fall of temperature with height ceases. 
As the velocity of sound in hydrogen is much greater than that 
in nitrogen or oxygen, it follows from this, that at very great heights 
the velocity of sound increases so much that the sound rays are 
curved towards the earth. This is illustrated by fig. 8, which has 
been deduced from suppositions to be mentioned afterwards. 
Von peM Borne’s hypotheses differ only quantitatively from these 
— the entirely smooth character of the sound rays is quite in har- 
mony with his reasoning. Therefore it is really a pity that the words 
“total reflection” have not been avoided, and that he and his sup- 
porters, especially WeGeNer, have repeatedly spoken of “reflection 
against the hydrogen atmosphere’. Many writers, therefore, thought 
themselves entitled to dispose of this theory simply by drawing 
attention to the gradual character of the change in density. Our 
figure 8 shows that, notwithstanding that gradualness, a sharp limit 
is produced, within which no sound can be observed, unless parti- 
cular circumstances cause an entirely unusual course of the rays of 
little elevation. 
Without for the moment going into the details of the numerical 
results, which are obtained according to different hypotheses about 
the constitution of the atmosphere, we indicate here the principal 
features which the silent region ought toshow according to thisexplanation. 
1. At the outer limit of the silent region a comparatively large 
intensity of sound must be expected. The intensity of the sound 
within the surface between A and 4 ought to be much greater than 
that between C and D. 
