Small graphite crucibles with covers, as shown in section, in Fig. 2, serve to guard against losing the ball, to handle it by when hot, and to protect it against loss of heat during transmission from the fire to the pyrometer. To guard against overturning the crucibles, moulded firebrick should be provided to receive them, two crucibles being put into one brick, in the same exposure, whenever great accuracy is desired, each serving as a check on the other, and their mean being likely to be more nearly correct than either one if they differ. The firebrick cover is occasionally useful to retard cooling, if, by reason of local obstructions, some little delay is unavoidable in transferring the balls from the fire to the water of the pyrometer. With convenient arrangements, this may be done in three seconds. After observing the temperature of the water, make ready for the immersion of the heat carrier by raising the agitator until a space of only about 1.5 of an inch is left between its rim and the cover. An instant before putting in the heat carrier--"pouring" it from the crucible--lift the cover and agitator both together, so that the rim of the latter is level with the sloping top of the instrument. The agitator then receives the hot ball without shock, and no harm is done. If the ball goes below the agitator, it is likely to injure the bottom of the cup. If, on taking the temperature of the water before the immersion of the heat carrier, any change is observed, either rising or falling, the direction and rate of such change, and the exact interval of time between the last recorded observation and the immersion, should be noted, in order to determine the exact temperature of the water at the instant of immersion. The temperature of the water will continue to rise as long as the heat carrier gives out heat faster than the cell loses it. The rise will grow gradually slower until it ceases, and the maximum can be very accurately determined. Examples of the mode of using the tables, and of determining the true temperature of the heat carrier at the instant of immersion from the observations with the instrument, are given in the table on pages 170 and 171 of this Journal for September. A method of using the tables, by which a closer approximation to the true temperature may be reached, will be pointed out in a subsequent article. [Illustration: Fig. 2.] DETERMINATION OF THE CALORIFIC CAPACITY OF THE METALS OF THE PYROMETER, in terms of water, i.e., in British thermal units.