Now this is diminished by the volume of the internal staying, and, although it was supposed that this would be inconsiderable, a correction was made for this. The arrangement of these stays caused the quantities by which the capacity was to be corrected, to be of two kinds. First, a constant correction for every depth, consisting of the volume of the perpendicular parts, the upright sheet-iron stays and rods, which amounted to 1379"62 cu. in. ~-"795 cu. ft. ~ 5"94 gallons; and second, a part, consisting of the angle iron two inches high, running around the bottom edge of the tank, and the angle iron and horizontal stay rods which were found to be a.t a height of 22~ inches above the bottom. This amounted to 919 cu. in. ~---"532 cu. ft.
"532 and--~ (~ "266)X 7"4805 ~ 1"9898 gallons per inch of height, or 1'9898 X 4 ~ 7"9529 gallons in all. The ultimate correction was therefore 13"906 gallons, and the real capacity 2502"3836 gallons.
This process was gone through four times, and the capacities found were as follows :--1st trial, 2453"23 gallons; 2d trial, 2519'44 gallons ; 3d trial, 2465"57 gallons ; and 4th trial, 2502"38 gallons, the variations depending on very slight differences indeed, in the measurements and decimal places.
Incidentally, the weight of the water contained in the tender when full may be here stated. It amounted to 336"3799 cu. ft. :~ 62.4 ~---20,990 lbs. or 10'495 tons. Now, in order to calculate columns 16 and 17, the cubic feet of water corresponding to each inch of height of tank were obtained. Multiplying the area 94"1743 sq. ft. by "0833 ft. (~ 1 inch)we