The temperature he selected as the starting point was 70 degrees Reaumur=190 degrees Fahr. He preferred this to the boiling point, as he could obtain it accurately by first plunging the whole apparatus duly charged into nearly boiling water, then allowing it to fall to 70 degrees, and immediately plunging it into a mixture of pounded ice and water, where, by the aid of a little agitation, it remained steadily at the freezing point. The following table shows the results with the substances therein mentioned, the quantity in each case being 16 grains: ============+===========+=============+============ |RAW SILK AS| | HEAT LOST. |SPUN BY THE|SHEEP'S WOOL.|COTTON WOOL. | WORM. | | ------------+-----------+-------------+------------ 70° | -- | -- | -- 60 | 94´´ | 79´´ | 83´´ 50 | 110 | 95 | 95 40 | 133 | 118 | 117 30 | 185 | 162 | 152 20 | 273 | 238 | 221 10 | 489 | 426 | 378 ------------+-----------+-------------+------------ Total times | 1,284 | 1,118 | 1,046 ------------+-----------+-------------+------------ The clothing value or "warmth" of these substances as thus applied varies directly with these figures, representing the passage of heat or their "non-conducting" power. The experiments described were directed to the determination of the relative power of different materials. (We have only quoted the substances most generally used for clothing purposes.) These were followed by another series upon certain given materials differently arranged, or in different conditions of density. The same "passage thermometer" was used. The question to be determined was, whether the protecting power of the substances used in the previous experiments was due to the non-conduction of the material of those substances themselves, or whether the air imprisoned between their fibres was an important factor. If the silk, wool and cotton did all the obstructive work independently of the air, then the amount of obstruction should vary with the quantity of fibre. As, in the experiments already described, the fibres were loosely arranged round the bulk of the thermometer, it was easy to increase their quantity by packing them more closely, and yet retaining the other conditions of bulk, etc., the same. It was evident from the results, with 16 grains, 32 grains, 64 grains, that the protective power does not increase nearly in the same proportion as the quantity of material when packed in the same space. Had such been the case, and the rate of cooling proportionately retarded, the experiment (with eider down from the duck) with 64 grains should have been 1,304x4 5,216 seconds, instead of only 1,615 seconds.