68 
PHYSICS: C. BARUS 
Proc. N. a. S. 
180°) is more regular. The curve lies in higher values of 5. There is 
less repetition of sectional groups. Nevertheless the same number of six 
maxima occur ; i.e., each max. in figure 1 is reproduced and often accentuated 
in figure 2. The former minimum of 7 = 40 has shifted to a minimum 
somewhat above 20 cm. 
The third survey in y was made with the pipe again at 2 = 40 cm. 
above the origin and with the resonator on the table (in azimuth 180°) 
at y = 40 cm. The graph, fig. 3, for the plate reflector normal to y and 
advancing along it shows the same number of maxima as heretofore, with 
the sixth soon to vanish. The feature of the curve is the enormously 
developed second maximum, in contrast with the first, or the others. 
We are no doubt trenching on the case of favorable incidence heretofore 
developed. The response was always stronger to the weak notes and they 
were used throughout, the difference being often over 10 scale parts, 5. 
Finally, figure 4 gives the corresponding acoustic distribution evoked 
by the advancing reflector with the pipe left in its former position, but with 
the resonator on the table at ;v = 65 cm. in azimuth 180°. The first 
maximum is now relatively high, while the former huge second maximum 
has flattened and the fifth and sixth maxima have vanished behind the wall. 
Weak notes were essential, as the strong notes gave relatively as little 
deflection, as before. As a whole, the present work adduces strikingly 
clear-cut evidence of the interferences to be identified in the following 
treatment; but it fails, as yet, to shed definite light on the anomalous 
distributions of intensity, s. 
These maximum ordinates, 5^, for path-differences X = 0, X, 2X, 3X, 
are constructed for the corresponding resonator positions, yr in figure 5. 
Compared with the regularly diminishing initial maximum, Xo the strongly 
harmonic 5-distribution for the X maximum, together with its decreasing 
amplitude are well brought out. Like Xo, 2X is but sparingly harmonic, 
while the odd 3X is again markedly so. The graphs, fig 5, also contain the 
relation of the reflector and resonator positions in y, for Xi crests. The 
broken line should probably be straight, at an angle of about 50° to the 
vertical. 
2. Discussion of the Preceding Reflections. — In the graphs 1 to 4, there are 
three influences at work on the resonator. In the first place the direct 
ray from the pipe is equivalent to a reflection from the table, a locus of 
re enforcement. A half wave-length has been lost. Then there are the 
two reflections, also with loss of X/2, from the board normal to y and the 
wall W Sit y — — 130 cm. This may be neglected, because it is always 
kept relatively very distant in the plan of the experiments. We thus 
have to compound the ray from the image of the source with the direct 
ray. The computation would be cumbersome; but the construction in 
the graph itself is not so and adequately correct. In figure 1, if the promi- 
