210 Proceedings of Indiana Academy of Science. 
ported the equation only as far as concerns variation of speed with 
diameter of pipe, and were in disagreement as to the effect of pitch. 
There is therefore no consensus of opinion on any of the points con- 
cerning the velocity of sound waves in tubes. It will be noted, however, 
that in no case has an observer claimed a greater speed in pipes than in 
the open. The writer has obtained such results. 
Figure 2 shows the general arrangement of the apparatus used in 
this experiment. The reader is referred to earlier papers' for a more 
detailed description and explanation. It will suffice here to say that 
two spark gaps S and I are in series and connected—through two vari- 
able gaps G, G—to the terminals of a powerful electrostatic machine. 
When the gaps G and G are shortened a discharge passes through the 
commutator C to the cireuit including the sound gap S and the illumin- 
ating gap I, the latter spark being retarded slightly by a variable ca- 
pacity K. By varying the capacity K and the length of the gap I, the 
light from the spark at I can be adjusted to cast a shadow on a photo- 
graphic plate P of the sound wave produced by the spark at S. 
Plate I shows such a wave. The sound spark was produced just 
behind the center of the circular screen (a hard rubber disk) D, the 
screen being used merely to prevent fogging the dry plate by the light 
of the sound spark. T is an end-on shadow of a portion of a piece of 
brass tubing 3 cm. in diameter and 5 em. long. The projecting arms 
are four pieces of brass tubing, respectively 0.25 cm., 0.48 em., 0.8 cm., 
and 1.15 em. in internal diameter, each of them 2.4 cm. long. They were 
soldered radially in holes whose diameters corresponded respectively to 
the outside diameters of the tubes. Almost half of the side wall of the 
supporting tube was then cut away, to permit the sound wave to travel 
out on one side in free air, while on the other side the wave was arrested 
except for the portions passing through the four radial tubes. The 
sound gap was placed as accurately as possible at the center of the 
supporting tube and the point of intersection of the axes of the radial 
tubes. 
In order to show at a glance just what has happened with the posi- 
tion of the sound gap as center I have drawn a broken line circle. To 
avoid confusion I have drawn the circle C just outside the main wave W. 
It will be noted that the waves through the tubes lie well without the 
circle, showing that the waves in the tube traveled more rapidly than 
the wave in free air, and that apparently the velocities in the several 
tubes were the same, although the tube diameters were in the approxi- 
mate ratios of 1, 2, 3 and 5. 
On the negative from which Plate I is a reduced print the waves 
1 Physical Review, Vol. 35, p. 373, 1912. Also Proceedings Indiana Academy of 
Science, p. 305, 1915. 
