RADIATION AND ABSORPTION. 81 
The only experiments that I have yet been able to make permit 
us to arrive however at a certain approximation ; they give 
— 142° 
for the temperature of space, and I do not think that this value 
can be far from the truth; it corresponds to 6=0°9. Thus we 
see, as the definitive result of these researches, that the sun im- 
parts to the earth a quantity of heat 1°7633 per minute and per 
square centimetre, that with a clear sky the atmosphere absorbs 
about four-tenths of this heat and of that of space; that it 
absorbs nine-tenths of the heat emitted by the earth, and that 
the temperature of space at the present period is 142° below 
zero. We can scarcely draw attention too strongly to the im- 
portant part which the inequality of the absorbing powers of the 
atmospheric air exercises on the whole of the terrestrial phe- 
nomena, and consequently to the care necessary to determine 
them with accuracy. For this purpose doubtless other appa- 
ratus and other methods of experimenting will be soon in- 
vented, by means of which it will be possible to separate at 
every instant the complex influences of the radiation of space 
and the atmospheric radiation. If at the present time the dif- 
ferent regions of the sky which pass successively to the zenith 
appear to us to transmit equal quantities of heat, it is very pro- 
bable that this only results from the imperfection of our appa- 
ratus: we perceive such differences in the nature, distance, num- 
ber and grouping of the stars in the depths of space, that it is 
impossible to admit that the portion of the heavens, incessantly 
changing, which is above the horizon, resembles constantly the 
portion that is below; and consequently it is impossible that all 
the hemispheres which we can conceive in the celestial vault 
transmit really to the earth the same quantity of heat. It is espe- 
cially in the equatorial zone that we must first seek to appre- 
ciate these differences, because they must doubtless appear there 
greater, more regular and more easy of observation. 
23. The following table contains the result of experiments 
made with the actinometer: the progressive depression of the 
zenithal temperature will be remarked; the last column of this 
table contains the mean temperature z" of the atmospheric co- 
lumn at Paris, corresponding to each observation, and calculated 
by the formula (4) of the zenithal temperature in which this 
quantity 7 remains alone unknown. 
VOL. IV. PART XIII. G 
