98 
PHYSICS: C. BARUS 
Proc. N. a. S. 
the same strongly marked harmonic character of the reflection graphs 
of the preceding paper. Cf. § 3. 
3. Pipe Displaced in a Given Level along y. — Figures G-10 are constructed 
on the same plan as the preceding, the ordinates s, denoting fringe de- 
flections or the nodal intensity, encountered by the mouth of the res- 
onator at y^i. If the ordinates be also taken as Zp, the pipe position, 
P, etc., may also be indicated on the graphs. T, T' show the edges of 
the table. In figure 6, where the pipe is 40 cm. above the origin, the graph 
is highly detailed with crests mostty 30 cm. apart. The hyperbolas 
for reflections from the wall at y = 174: cm. and be5^ond T' have been 
sketched in at 5X and 6X of path difference, together with the corre- 
sponding acoustic rays. So also those for reflection from the wall at 
y = — 130 cm. (beyond T). The last agrees badly with the crest observed 
at about —10 cm. The alluring possibility of an intersection of the 
positive and negative hyperbolas resulting in very high crests does not 
work out. In figure 7 also the positive hyperbola 5X fairly corresponds 
to the observed crest, but not the negative one. In figures 8, 9, 10 the 
positive hyperbolas 3X, 4X, 5X, also, fail of coincidence with the observed 
crests. These in fact lie nearly between them, or half a wave-length of 
path difference beyond. Thus it seems improbable that the maxima and 
minima can be explained as resulting from the reflection at these walls, 
while all other walls and the ceiling are far more distant. 
In figures 6-10 one is often in doubt as to which of the crests belong 
together. Crests seem rather to separate and recombine. Taking the 
highest in the successive curves, figure 11 shows that the relation of pipe 
and resonator positions for maxima is here not simple as in figure 5. In 
fact the pipe approaches the high crest and finally overtakes it (figure 10). 
