Consider a ray of light falling on a surface from the Sun. A part of it is reflected; that is, is instantly thrown off again. By this part the body shines and makes its show in the world, but gets no good itself. Another part is absorbed; this alone goes to heat the body. Now if the visible rays were all that emanated from the Sun, it would be strictly true, and a pretty paradox for believers in the efficacy of distance, that what heated the planet was precisely what seemed not to do so. Unfortunately there are also invisible rays, and these, too, are in part reflected and in part absorbed, and their ratio is different from that of the visible ones. To appreciate them, Langley invented the bolometer, in which heat falling on a strip of metal produces a current of electricity registered by a galvanometer. By thus recording the heat received at different parts of the spectrum and at different heights in our atmosphere, he was able to find how much the air cut off. Very has since determined this still more accurately. By thus determining the depletion in the invisible part of the spectrum joined to what astronomy tells us of the loss in the visible part, we have a value for the whole amount. By knowing, then, the immediate brightness of a planet and approximately the amount of atmosphere it owns, we are enabled to judge how much heat it actually receives. This proves to be, in the case of Mars, more than twice as much as distance alone would lead us to infer. The second question is how much of this it retains. The temperature of a body at any moment is the balance struck between what it receives and what it radiates. If it gets rid of a great deal of its income, it will clearly be less hot than if it is miserly retentive. To find how much it radiates we may take the difference in temperature between sunset and sunrise, since during this interval the Earth receives no heat from the Sun. In the same way the efficacy of different atmospheric blankets may be judged. Thus the Earth parts with nine centigrade degrees’ worth of its store on clear nights, and only four degrees’ worth on cloudy ones, before morning. This is at sea-level. By going up a high mountain we get another set of depletions, and from this a relative scale for different atmospheric blankets. This is the principle, and we only have to fill out the skeleton of theory with appropriate numbers to find how warm the body is.