Vol. 8, 1922 
PHYSICS: C. BARUS 
97 
sively in steps of 20 cm. from z = 20 cm. to z = 80. For each pipe level 
the distribution of nodal intensity along the y axis was explored by mov- 
ing R along it (with observations taken every Ay^ = 10 cm. apart). In 
figure I, Zp = 20 cm., there is probably some distortion at the origin, 
owing to the wind from the near pipe, otherwise the first crest would 
be in negative y. From the glancing incidences, this graph has less 
detail than figure 2 {zp = 40 cm.), which preserves the same general 
character but is lower in 5 and shrunk in the y direction. In figure 3, 
0 = 60 cm., the topography is totally changed. There is now a trough 
20 40 60 80 ■ m if20 
40 80 
near the origin and (on compounding) but one marked crest. Figure 
4, Zp = 80 cm., is redeveloping a crest at the origin, but otherwise pre- 
serves the former outlines. In this figure and the following, the mouth 
of the resonator was used to define yj^ so that the two curves are practically 
coincident. 
If the highest crests correspond to like path differences and there- 
fore belong together, their position yj^, Zp in figure 5 is seen to progress 
roughly along a straight line y = 1.25 2; — 40, the slope dy/dz corre- 
sponding to an angle of incidence of o: = 51°. The minima determine 
parallel lines, the vertical distance apart of consecutive lines being here 
Azp = 24 cm. or about half a wave-length. 
Furthermore, the graphs, as a whole, are alternately high and low in 
intensity, 5. If the intensity of the highest crest be expressed in terms 
of the pipe elevation 0^, the resulting graph, figure 2' (scale V2), presents 
