There is a fourth boiler, designed and built for testing purposes by the Quartermaster’s Department. This is a tubular boiler designed on the lines of a house-heating boiler, but for use as a calorimeter to determine the relative heat value of different fuels reduced to the basis of a standard cord of oak wood. A series of research tests on the processes of combustion is being conducted in Building No. 13, by Mr. Henry Kreisinger. These tests are being made chiefly in a long combustion chamber (Figs. 16 and 17, and Figs. 1 and 2, Plate XVIII), which is fed with coal from a Murphy mechanical stoker, and discharges the hot gases at the rear end of the combustion chamber, into the hand-fired Heine boiler. The walls and roof of this chamber are double; the inner wall is 9 in. thick, of fire-brick; the outer one is 8 in. thick, and is faced with red pressed brick. Between the walls of the sides there is a 2-in. air space, and between them on the roof a 1-in. layer of asbestos paste is placed. The inner walls and roof have three special slip-joints, to allow for expansion. The floor is of concrete, protected by a 1½-in. layer of asbestos board, which in turn is covered by a 3-in. layer of earth; on top of this earth there is a 4-in. layer of fire-brick (not shown in the drawings). [Illustration: Fig. 16. CROSS-SECTIONS OF CHAMBER AND OF FURNACE, LONG COMBUSTION CHAMBER] Inasmuch as one of the first problems to be attacked will be the determination of the length of travel and the time required to complete combustion in a flame in which the lines of stream flow are nearly parallel, great care was taken to make the inner surfaces of the tunnel smooth, and all corners and hollows are rounded out in the direction of travel of the gases. Provision is made, by large peep-holes in the sides, and by smaller sampling holes in the top, for observing the fuel bed at several points and also the flame at 5-ft. intervals along the tunnel. Temperatures and gas samples are taken simultaneously at a number of points through these holes, so as to determine, if possible, the progress of combustion (Fig. 1, Plate XVIII). About twenty thermo-couples are embedded in the walls, roof, and floor, some within 1 in. of the inside edge of the tunnel walls, and some in the red pressed brick near the outer surface, the object of which is to procure data on heat conduction through well-built brick walls[16] (Fig. 2, Plate XVIII). In order to minimize the leakage of air through the brickwork, the furnace and tunnel are kept as nearly as possible at atmospheric pressure by the combined use of pressure and exhausting fans. Nevertheless, the leakage is determined periodically as accurately as possible.