658 
PHYSICS: A. G. WEBSTER 
Proc. N. a. S. 
this the calculations were made by my assistant, Dr. E. A. Harrington, 
who has carried out all the experimental work except the determination 
of the travel of the shot. In the figures (p. 649), curve 1 represents p,t for 
the shot used, while 2 is a curve of a reduced charge of 7/10 normal. The 
curves were then graphically integrated twice by counting the squares. 
Curves 3 and 4 show the first and second integrals pdt and J*dtJ^ pdt, 
respectively. The travel of the shot was determined by observing the 
time of contact of the bullet with the end of a coil of fine wire' forced 
down the barrel to a measured point, an oscillograph of high frequency 
being used, and the time being observed on one of the revolving drums 
previously described. This was done about a year ago by my then assis- 
tant, Mr. H. C. Parker, and the result, shown in curve 5, is the average 
of a good many shots. The exact agreement of the two curves 4, 5 shows 
the propriety of the assumption about the resistance, and determines b 
as (5 — 6)/5 = 0.894. In this case Ro = 0. We can also neglect po 
so that F = 0. 
The muzzle velocity and the time of the shot leaving the barrel were 
determined by the method described in these Proceedings, April, 1920, 
developed in this laboratory by Thompson, Hickman, and Riffolt. The 
constants of the gun were as follows : 
As for the constants pf the powder, the density was determined by Dr. 
Harrington in the pyknometer as 1.58 gm./cm. No attempt was made 
to determine rj experimentally, that being best done in the bomb, but the 
value 7] = 0.95 was assumed. The load was co = 3.100 gm. To deter- 
mine /, the "force" of the powder (which has the dimensions of energy/ 
mass) two methods were used, which agreed very well. If it is assumed 
that' the point of inflexion on the observed pressure curve marks the end 
of combustion, z = 1, / is determined from equation (19). If, on the other 
hand, equation (19) is used with observed values of p, %, and u, it will 
give values of z increasing to a maximum, and then diminishing. This 
maximum should be equal to unity, and after a few trials a value of / 
can be found that makes it so. As stated, this method agrees with the 
other. We obtain / = 1.087 X 10^° cm.Vsec.^, corresponding very well 
with the values for the French powders and cordite. (The values given 
by Charbonnier and Sugot involve gravitation units.) We thus obtain 
If we take k = 1.2, and X = 0.2, we have G = 0.333. Also we find 
that g (eq. 20) varies by about 16%, nearly as a linear function 
of X or s:. 
Volume of chamber, c = 4.18 cm. 
Area of bore, S = .456 cm. 
B = 0.716 cm.Vgm., 
D = 0.32 cm.Vgm., 
C = 0.147 cm.Vgm., 
E = 0.394 cm.Vgm. 
