174 PHYSICS: THOMPSON, HICKMAN, RIFFOLT Proc. N. A. S. 
velocities ranging perhaps from 1200 feet per second up to 3500 or 4000 feet 
per second can be made at each end of the tube. As long as the velocity of 
the projectile is greater than that of sound, no interference with the meas- 
urements will result from a disturbance reflected from the walls of the tube. 
m zoo 300 m 500 eoo 100 300 m mo 1100 /200 mo m /300 m /m 
vf/oc/n /A/ mfRS PfR sfcow 
FIG. 4 
Figure 4 is a graphical representation of the resistance as a function of 
the velocity as determined from the measurements made by at least 
five European observers, their data being averaged by the Italian ballis- 
tician, General Siacci. The retardation appears to be a linear function 
beyond the velocity of sound. Multiplying the datum line of these curves 
by different factors depending upon the density of the air gives the other 
curves for various heights. These were used in trajectory computations 
on long range guns of the type brought out by the Germans to shell Paris. ^ 
The experiments outlined will be adequate to make a precise determina- 
tion of the values for the resistance for small projectiles over a consider- 
able part of the usable portions of this graph for the modern rifles of high 
velocity. The fact that this problem is not the exact equivalent hydro- 
dynamically of the one for very large projectiles does not materially lessen 
the value or interest of the results of such an investigation with small 
projectiles. 
An important problem is the accurate determination of the effect of 
varying the form of the projectile. The question of ultimate or optimum 
form for a given use should be experimentally determinate. The records 
show not only the velocity but the manner in which the projectile is 
travelling. Also, it is likely that the precision and the complete control 
of the conditions of experiment possible would lead to satisfactory results 
