Carbonic oxide and carbonic acid are pervious to a vast majority of the rays of radiant heat. When the cylinder was filled with carbonic oxide gas and so divided, by moving the internal plate of rock-salt, that a stratum of the gas 8 inches long was next to the source of heat, and that 41·4 inches long farthest from it, the 8 inches of gas intercepted 6 per cent. of the whole radiation. But when the plate of rock-salt was moved till the column 41·4 inches long was next to the source of heat, and that of 8 inches farthest from that source, or behind the long one, the absorption of the 8 inches was sensibly zero. In like manner eight inches of carbonic acid gas when in front of a column of 41·4 inches of the same gas absorbed 6-1/4 per cent. of the whole radiation, while placed behind that column the effect was nearly zero. The reason is that when the 8 inch stratum is in front, it stops the main portion of the rays which give it its thermal colours,[7] while placed behind these same rays have been almost wholly withdrawn, and to the remaining 94 per cent. of the radiation the gases are sensibly permeable. It is inferred from an extension of this reasoning that the sum of the absorptions of the two chambers taken separately must always be greater than the absorption effected by a single column of the gas of a length equal to the sum of the two chambers; this conclusion is illustrated in a striking manner by the experiments. It is also found that when the mean of the sums of the absorptions is divided by the absorption of the sum, the quotient is sensibly the same for all gases. It may farther be inferred that the sum of the absorptions must diminish and approximate to the absorption of the sum as the two chambers become more unequal in length, and that the sum of the absorptions of the two chambers is a maximum when the medial plate of rock-salt divides the long tube into two equal parts. When air enters an exhausted tube it is heated dynamically by the collision of its particles on the sides of the tube as it rushes in to fill the vacuum; and when the tube is exhausted again by the air pump, chilling is produced by the application of a portion of the heat of the air to generate vis viva. This dynamic principle occurred in some of the experiments, and was dexterously adopted and applied to the solution of a striking and unprecedented problem: ‘To determine the radiation and absorption of gases and vapours _without any source of heat external to the gaseous body itself_.’