78 M. POUILLET ON SOLAR HEAT, 
The numerous results of Mr. Wells, of Mr. Daniell, and of all 
the other natural philosophers who have made experiments on 
nocturnal radiation, not only prove that a thermometer exposed 
on the ground during the night, in an open place, is cooled 6°, 
7°, or even 8° below the ambient temperature; they also prove 
that this phenomenon is reproduced, almost with the same in- 
tensity, in the coldest months of the year, that is to say in 
January and February, when the temperature of the air has 
fallen many degrees below zero. Thus Wilson observed a dif- 
ference of nearly 9° between the temperature of the air and that 
of the surface of the snow; Scoresby and Captain Parry have 
observed analogous depressions in the Polar regions, when the 
temperature of the air was more than 20° below zero. 
If we now consider that the heating power which the stratum 
of air exerts by its contact with the thermometer of the ground, 
which is colder than it, is nearly the same, whether it be at 10° 
above zero or 10° below, it results that the cooling power which 
maintains this thermometer at —-18° in the second case, has also 
the same energy as the cooling power which maintains it at + 2° 
in the first case; and, as this cooling power depends on the 
temperature of space, it follows also that the temperature of 
space is much lower than —18°; for if it were only —30° or 
—40°, the thermometer which is at —18° whilst the air is at 
— 10° would be too near for the heat of space to be able to keep 
it at the same depression below the air as the thermometer which 
is at + 2° whilst the air is at 10°. What has perhaps hitherto 
prevented our arriving at this conclusion, is that in the expla- 
nations which have been given of the nocturnal radiation, a pecu- 
liar cooling power has been generally attributed to the upper 
strata of the atmosphere, which are known to be very cold, for- 
getting in some measure that, cold as they are, it is nevertheless 
heat which they transmit, and that this heat is added to that of 
space to augment its effects. 
The results which I have obtained by means of the actinome- 
ter are therefore in accordance with all known facts ; it was per- 
haps essential to make this remark, in order to show that, if the 
consequences at which we shall arrive are in some points con- 
trary to received opinions, this results from the nature of things 
rather than from the inaccuracy of the experiments. 
22. Considering the equation (4.) as an equation of condition, 
which must always be satisfied for every value of the zenithal 
temperature given by experiment, I have been able to determine 
