The results were not in all cases consistent with each other, but they indicate that the property is dependent on the texture of the substance rather than on the kind of material, or—as concerns non-luminous heat its color.
The most recent experiments are those of Dr. Krieger, some results of which are cited by Dr. Pettenkofer. He observed the rate of cooling of a sheet-iron cylinder filled with hot water and covered by turns with different cloths. Wrapping it with successive envelopes of wool, buckskin, silk, cotton, and linen, and observing regularly the diminution of temperature in a given time, he found the differences insignificant, not exceeding one or two per cent. The color of the materials did not cause the results to vary any more. It appears, then, that in a dark heat the emissive power and the absorbing power, which is correlative with it, vary but little between one kind of cloth and another. The case is different when we have to do with luminous heat, or the solar rays. With envelopes of linen, cotton, flannel, and silk, M. Krieger observed that the absorption of solar heat increased in the proportions indicated by the numbers 90, 100, 102, and 108. The influence of color was much greater: with cotton goods of different hues he found the numbers to be white, 100; straw-color, 102; yellow, 140; bright green, 155; dark green, 168; Turkey red, 165; bright blue, 198; black, 208. These facts explain why in the hot sun a black coat is warmer than a white one, while the difference disappears in the shade. The influence of colors on the absorbing powers of surfaces had already been made clear by the researches of Leslie and Melloni.
To form an estimate of the part which the conductibility proper of the different materials plays in these phenomena, M. Krieger inquired how much the loss of caloric was diminished when the cylinder was covered with double layers of the same cloths. The doubling of the satin, cotton cloth, and fine linen diminished the loss only by from three to six per cent, while doubling the envelopes of buckskin, flannel, and woolen cloth diminished it by ten, twenty, and even thirty per cent. It is clear from these experiments that the resistance offered by cloths to the passage of heat depends much less on the conductibility of the fibers that form their substance than on the thickness, the volume, and the texture of the tissues. This can also be shown in observing the cooling of a cylinder covered with wadding, which is forty per cent more rapid when the wadding is strongly. So a dressing-gown lined with wadding and a flannel waistcoat are warmer when we first put them on than after they have been worn for some time. The packing which the filaments undergo with use renders the cloth more permeable to heat. Although doubling the envelope has little influence when both layers are stretched tight over the cylinder, it is not the same when a slight space is left between them; then the cooling is retarded thirty or thirty-five per cent by