First. Weigh the cup, or cell, the lower plate of the cover and the metallic portion of the agitator, and compute their heat-capacity by the specific heat of the respective metals. Compute also the heat capacity of the thermometer; or, if it be long, of so much of it as is found to share nearly the temperature of the immersed portion. The result will be a minimum--indeed, in so small a vessel the inevitable loss by conduction and radiation will amount to more than one-third as much as the simple heat capacity of the metals.[1] The total must be ascertained by an application of the method of mixture. Ascertain the temperature of the interior of the instrument simply; pour in quickly but carefully a known quantity of water, say about two pounds, of known temperature, say about 100° F., and ascertain the temperature as soon after pouring as mixing can be properly performed. But a correction is necessary for loss of heat in the act of pouring. To ascertain the amount of this correction prepare a bath of tepid water, and bring all parts of the instrument--outside, inside, and interior portions, together with the vessel to pour from--exactly to one common, carefully ascertained temperature. Now take two pounds of the water and pour it into the cell in the same manner as before. Exposure of so thin a stream on two surfaces to the air of the room will produce a certain degree of refrigeration in the water, which is supposed to be warmer than the air, say at about 160° F. This effect will be due to conduction, by contact with the air, to radiation, and to evaporation; and by so much the refrigeration observed in mixing is to be diminished. [Footnote 1: In our case the heat-capacity, thermometer included, was 0.0757; total, 0.1053; radiation, etc., 0.0296. Respectively, 71.9 per cent, and 28.1 per cent. of the total.] Four experiments, carefully conducted, gave the following results: Loss of temperature by pouring at 170° F., 0.81°, 0.86°, 1.00°, and 1.07° F.; mean, 0.935° F. The following are values of the calorific capacity of my pyrometers, that is, of those parts of each which share directly the temperature of the inclosed water, including the thermometer to be used with the instrument, and the heat communicated to the eider-down and otherwise lost during an observation, expressed in decimals of a British thermal unit, or in decimals of a pound of cold water: 0.1048, 0.1052, 0.1077, 0.1008, 0.1028, and 0.1104. Mean 0.1053 = 0 lb. 1 oz. 11 drms. Add water 1.8947 = 1 " 14 " 4 " ------ - -- -- 2.0000 = 2 " 0 " 0 " This was the value used. The instrument, being put on delicate coin scales and counterbalanced, weights equal to 1.8947 lb. avoirdupois = 1 lb. 14 oz. 5 drms., were added to the counterbalancing weights, and cold water was poured in until the scales again balanced.