Vol.. 8, 1922 
PHYSICS: C. BARUS 
121 
its two marked crests. Somewhat less certainly the resemblance in W-E 
lines may be detected. As in all the N-S lines the central trough lies for 
some reason to the right of the pipe normal. The maxima are far apart 
apparently. 
The resonator was finally transfered to a level, X/2 = 24 cm. above the 
table and moved around the pipe axis, its mouth lying on successive cir- 
cumferences of radii r, with the results given in figure 5. The maxima of 
all the curves are here curiously flattened; but there is still a gradual 
transfer of the chief crest from S to W, as r increases. Probably inter- 
mediate observations would have modified the curves. Since X/2 de- 
termines a nodal plane normally below the pipe, the intensities are all 
relatively large, but they fall off as a whole so rapidly with r, that there 
is no intersection and all curves may be given in a single diagram. 
The graphs showing the intensity, s, along the S-N and W-E directions 
in figure 8, exhibit a total change of form. The two crests have seemingly 
all but coalesced near the pipe normal. Owing to an accident which some- 
what dislocated the fringes, the indentation in the middle of the S-N 
curve and the bulge toward the right in the other may need modification 
but this is of little consequence here, since the plane at a level oi z — 24; 
above the table contains a single marked crest, only. 
2. Vertical Survey. — It is now of interest to bring the values of intensity 
together when the resonator moves up from the table on the same vertical. 
As constructed in figure 9, with 5 laid off horizontally to the right, 
vertically and with the origin of each tripoint s-curve at the x position of 
the vertical to which it belongs, these vertical distributions of nodal in- 
tensity strike the eye. The intensities 5 on the table for the same x posi- 
tions are also sketched in. Under the pipe at z^ = 50 cm., the nodal 
surface at the table rapidly decreases in s-intensity to the antinode at 
•2^R = ^/4; ^ then rapidly increases again to the more pronounced nodal 
surface at 2;r = X/2, nearer the pipe. Toward the south however the 
distinction between node and antinode is not regularly sustained: at x 
= — 20 cm. the antinode has gained in strength and at x = — 30 cm. 
the strength of nodes and antinodes does not differ much. Beyond this 
{x = — 40, — 50), however, nodes and antinodes are again sharply 
contrasted. Toward the north from the pipe normal (x = 0), these 
differences die out more gradually, until at x = 50 cm. the intensities are 
for some further reason, reversed. 
What is very surprising, however, is the observation that throughout 
the whole of the extent of a linear meter (^c = ± 50) symmetrically to the 
pipe normal, the nodal and antinodal planes at 0r = 0, X/4, X/2, remain 
parallel to the table, so far as can be seen. The only distinction within 
this stretch is the distribution of nodal intensity as exhibited in the con- 
tinuous curves sketched in, or separately in figures 6, 7 and 8. It would 
