At first a number of tests were run to calibrate the apparatus as a whole, all these preliminary tests being made on cheap, carefully inspected, uniform screenings from the same seam of the same mine near Pittsburg. Later tests will be run with other coals of various volatile contents and various distillation properties. It is anticipated that the progress of the tests may suggest changes in the construction or operation of this chamber. It is especially contemplated that the section of the chamber may be narrowed down by laying sand in the bottom and fire-brick thereon; also that baffle walls may be built into various portions of it, and that cooling surfaces with baffling may be introduced. In addition to variations in the tests, due to changes in construction in the combustion chamber, there will be variations in the fuels tested. Especial effort will be made to procure fuels ranging in volatile content from 15 to 27 and to 40%, and those high in tar and heavy hydro-carbons. It is also proposed to vary the conditions of testing by burning at high rates, such as at 15, 20, and 30 lb. per ft. of grate surface, and even higher. Records will be kept of the weight of coal fired and of each firing, of the weight of ash, etc.; samples of coal and of ash will be taken for chemical and physical analysis, as well as samples of the gas, and other essential data. These records will be studied in detail. [Illustration: Fig. 17. LONGITUDINAL SECTIONS OF LONG COMBUSTION CHAMBER] A series of heat-transmission tests undertaken two years ago, is being continued on the ground floor of Building No. 21, on modified apparatus reconstructed in the light of the earlier experiments by Mr. W. T. Ray. The purpose of the tests on this apparatus has been to determine some of the laws controlling the rate of transmission of heat from a hot gas to a liquid and _vice versa_, the two being on the opposite sides of a metal tube. It appears that four factors determine the rate of heat impartation from the gas to any small area of the metal[17]: [Footnote 17: The assumption is made that a metal tube free from scale will remain almost as cool as the water; actual measurements with thermo-couples have indicated the correctness of this assumption in the majority of cases.] (1).--The temperature difference between the body of the gas and the metal; (2).--The weight of the gas per cubic foot, which is proportional to the number of molecules in any unit of volume; (3).--The bodily velocity of the motion of the gas parallel to any small area under consideration; and (probably), (4).--The specific heat of the gas at constant pressure.