March 24, 1922] 



SCIENCE 



323 



pin hole and measured at a mercury U -gauge, 

 read by displacement interferometry. As the 

 pipes to be employed were to be of all kinds 

 and intensities, it did not seem worth while to 

 reduce the fringe-deflections to pressures. 

 These deflections will therefore be reported as 

 measured on an arbitrary scale s (.1 mm. col- 

 limator plate micrometer). The width of a 

 fringe was however on the average about 2 

 scale parts. Thus the corresponding pressure 

 increments p are readily found from 

 j3 := .00015 s millimeters of mercury. The 

 apparatus can be made more sensitive by en- 

 larging these small fringes; but the latter 

 would then usually be thrown out of the field 

 of the telescope and the screw micrometer be- 

 come necessary to restore them, which is irk- 

 some, particularly with fringe U -gauges. 



The pin-hole probe at one end of an eighth 

 inch pure rubber tube of any length (2 to 3 

 meters or more), at the other end of which is 

 the gauge, is at once available for introduction 

 anywhere. It fails however to give an appre- 

 ciable acoustic record except in the inside 

 sounding pipes. The plan of associating the 

 pin hole probe with a resonator, of either the 

 open or closed type, thus suggests itself. 



2. The closed pin hole resonator. The out- 

 standing trouble encountered heretofore was 

 ascribed to the organ pipe; i. e., to the con- 

 tinuity of notes lying very close together, but 

 to only one of which the resonator responds 

 effectively. A further difficulty was referable 

 to two ends of the open cylindrical resonator 

 then used, with a pin hole in the middle 

 within; for these ends being 22 cm. apart are 

 liable to lie in regions differing acoustically. 

 The advantage of the tube is the ease with 

 which it may be accurately tuned by mere 

 elongation and hence its sensitiveness. The 

 Helmholtz resonator, probably for this reason, 

 was found much less sensitive. Hence the 

 closed pin hole resonator qR, figure 2, suggests 

 itself, consisting of the cylindrical tube (1.9 

 cm. diam., effectively 22 cm. long) R, closed 

 by the snugly fitting cork c which carries the 

 pin hole probe qO. The pin hole, 0, is at the 

 base of the tube R, the quill tube q being con- 

 nected, as stated, by a length of gum rubber 

 tubing with the U -gauge. This resonator has 



but one mouth and thus tests acoustically a 

 single point, as it were, of the region, while 

 the tuning may be effected with nicety by 

 moving the cork c within R, or by elongation 

 at the mouth of R. The pin hole at must be 

 salient (i. e., carried by the conical end of the 

 quill tube q) and not on a reentrant or flat end. 

 Moreover, the diameter of the pin hole must 

 bear a certain relation to the size of the res- 

 onator R, to be found by trial. The construc- 

 tion of a sensitive pin hole resonator is ex- 

 tremely difficult; out of dozens of trials I 

 netted but one or two adequately sensitive 

 instruments. If of metal foil, it is liable to 

 change in the lapse of time. 



3. Survey between walls; y coordinate. The 

 two series of results (pipe in azimuth 180° at 

 a; = 2/ = 0, 2 ^ 40 cm., and resonator in azi- 

 muth 90° and 270°, respectively, at s = a; = 

 and y) are given in figure 3. The abscissas, 

 dy, are the distances between centers of pipe P 

 and resonator R. The y values of B (on the 

 table) are given in decimeters on the curves. 

 Insets show the orientation. The two curves 

 are pronouncedly harmonic from j/ = to the 

 wall, and they are ahnost exact inversions of 

 each other, crests in the one taking the place 

 of troughs in the other, throughout. More- 

 over the crests seem to lie in successive levels; 

 an initial high one {y = 10-20 cm.) ; an inter- 

 mediate lower one {y = 30-80 cm.); a still 

 lower one (100-140 em.) beyond; etc. It is not 

 feasible to go much beyond 160 cm. ; for with 

 the wall at 174 cm. the resonator pitch is be- 

 ginning to be modified. A curious result is 

 the initial maximum and minimum, the points 

 not being at y = 0, but beyond at y = 20 em. 

 The horizontal mean wave lengths Ay to be 

 obtained roughly from these curves are 

 respectively Ay = 34 cm. and Ay = 35.5 cm., 

 the range being from 30 to 40 em. The mean 

 d intervals, Adj., are about 30 cm. 



The precipitous descent of the graphs be- 

 tween y = 70 and 90 in one case and y = 90 

 to 100 in the other, makes less impression on 

 the wave intervals than would have been 

 anticipated and the mean wave length here, 

 X = 35 cm., does not differ from the earlier 

 cases of X =: 37 cm. by more than the ob- 

 servations of a single curve. The graphs, 



