The following are the details of the experiments: First series of experiments: Conduit consisting of cast or wrought iron pipes, joined by means of flanges, bolts, and gutta percha rings. D = 0.20 m. (8 in.); L = 4,600 m. (15,100 ft,); h= 26.77 m. (87 ft. 10 in.). 1st experiment: Q = 0.1860 cubic meter (6.57 cubic feet), at a pressure of ½(p1 + p), and a temperature of 22° Cent. (72° Fahr.); p1 = 5.60 atm., p =5.24 atm. Hence p1 - p = 0.36 atm.= 0.36 x 10,334 kilogrammes per square meter (2.116 lb. per square foot), whence we obtain b1=0.0001697. D'Aubuisson's formula would have given p1 - p = 0.626 atm.; and M. Arson's would have given p1 - p = 0.9316 atm. 2d experiment: Q = 0.1566 cubic meter (5.53 cubic feet), at a pressure of ½(p1 + p), and a temperature of 22° Cent. (72° Fahr.); p1 = 4.35 atm., p = 4.13 atm. Hence p1 - p = 0.22 atm. = 0.22 X 10,334 kilogrammes per square meter (2,116 lb. per square foot); whence we obtain b1 = 0.0001816. D'Aubuisson's formula would have given p1 - p = 0.347 atm; and M. Arson's would have given p1 - p = 0.5382 atm. 3d experiment: Q = 0.1495 cubic meter (5.28 cubic feet) at a pressure of ½(p1 + p) and a temperature 22° Cent. (72° Fahr.); p1 = 3.84 atm., p = 3.65 atm. Hence p1 - p = 0.19 atm. = 0.19 X 10,334 kilogrammes per square meter (2.116 lb. per square foot); whence we obtain B1 = 0.0001966. D'Aubuisson's formula would have given p1 - p = 0.284 atm., and M. Arson's would have given p1 - p = 0.4329 atm. Second series of experiments: Conduit composed of wrought-iron pipes, with joints as in the first experiments. D = 0.15 meter (6 in.), L - 0.522 meters (1,712 ft.), h = 3.04 meters (10 ft.) 1st experiments: Q = 0.2005 cubic meter (7.08 cubic feet), at a pressure of ½(p1 + p), and a temperature of 26.5° Cent. (80° Fahr.); p1 = 5.24 atm., p = 5.00 atm. Hence p1 - p = 0.24 atm. =0.24 x 10,334 kilogrammes per square meter (2,116 lb. per square foot); whence we obtain b1 = 0.3002275. 2nd experiment: Q = 0.1586 cubic meter (5.6 cubic feet), at a pressure of ½(p1 + p), and a temperature of 26.5° Cent. (80° Fahr.); p1 = 3.650 atm., p = 3.545 atm. Hence p1 - p = 0.105 atm. = 0.105 x 10,334 kilogrammes per square meter (2,116 lb. per square foot); whence we obtain b1 = 0.0002255. It is clear that these experiments give very small values for the coefficient. The divergence from the results which D'Aubuisson's formula would give is due to the fact that his formula was determined with very small pipes. It is probable that the coefficients corresponding to diameters of 0.15 meter (6 in.) and 0.20 meter (8 in.) for a substance as smooth as tin, would be still smaller respectively than the figures obtained above.