Hitherto the absorption and radiation of heat by the same thickness of different gases and vapours have been compared with each other, but in a recent series of experiments Mr. Tyndall has compared the action of different thicknesses of the same gas or vapour on radiant heat. The experiments extend from a thickness of 0·01 of an inch to that of 49·4 inches. The instrument employed for ascertaining the action of the smaller thickness was a horizontal hollow cylinder closed at one end by a plate of rock-salt. A second cylinder was fitted into this with its end also closed by a plate of rock-salt. This cylinder moved within the other like a piston, so that the two plates of rock-salt could be brought into flat contact with one another, or could be separated to any required distance, and the distance between the plates was measured by a vernier. At one end of the cylinder there was a source of constant heat, and the differential goniometer already described at the other. With this apparatus Mr. Tyndall found that olefiant gas maintains its great superiority over the other gases in absorptive power at all thicknesses. A layer of that gas not more than 0·01 of an inch thick intercepted about one per cent. of the total radiation. This great absorption corresponded to a deflection of 11° of the needle of the goniometer, and such was the delicacy of the apparatus that it would be possible to measure the action of a layer of this gas of less thickness than a sheet of writing paper. A layer of olefiant gas two inches thick intercepts nearly 30 per cent. of the entire radiation. A shell of olefiant gas two inches thick surrounding our globe would offer no appreciable hindrance to the solar rays in coming to the earth, but it would intercept, and in great part return, 30 per cent. of the terrestrial radiation; under such a canopy the surface of the earth would probably be raised to a stifling temperature. The apparatus for measuring the action of the greater thicknesses of gas was a hollow brass cylinder 49·4 inches long, closed at both ends by plates of rock-salt, and divided internally into two compartments or chambers by a third plate of rock-salt movable in the interior; the source of heat being at one end and the differential goniometer at the other.