i68 
PHYSICS: G. W. STEWART 
Proc. N. a. S. 
100 to 1200 d.v. were used. Purity of tone was approximated in a special 
apparatus, but the results were checked by tones from tuning forks and 
hence the conclusions are reliable. It was found that, if the phase differ- 
ences at the two ears were varied, intensities remaining equal and constant, 
the angular displacement of the phantom sound from the median plane 
was proportional to the phase difference at the ears. Inasmuch as the 
theoretical phase differences at the ears were known (Stewart, Physic. 
Rev., 4, 1914 (252) and Hartley, Ihid., 13, 1919 (373)), it became possible 
to ascertain whether or not the relationship between phase dift'erence and 
angular displacement was quantitatively the same as in the experiments 
just mentioned. When the comparison was made, it was found that there 
was a reasonably close quantitative agreement, i.e., the difference over 
practically all of the range was, on the average, less than 10%. In fact,, 
the correspondence between results in the theoretical case and in the ex- 
perimental one was so striking as to leave no doubt as to the fact that 
phase must be the most important factor in localization in the region 100 
to 1200 d.v. 
As can be readily understood, the higher the frequency the less that 
displacement from the median plane for which the phase difference is 
180°. For 1024 d.v., this displacement is 41°. But this is not the maxi- 
mum angle for which a source can be located by phase. This was shown 
by the existence of multiple phantoms or images. Thus the conclusion 
already stated seems worthy of full confidence as applicable for a dis- 
placement of less than 90° in front. This does not close the case with 
pure tones for there is yet to be ascertained what additional factors are 
of moment in locating a source not in the horizontal plane and not in front 
of the observer. 
That it is impossible to explain the difference of phase effect in terms of 
intensities, either by bone conduction or otherwise, is proved by the fact 
that the difference of phase effect was clearly present at 1024 d.v., with 
two observers who could not at this frequency get the intensity effect at 
all, i.e., the moving phantom source occasioned by altering intensity ratios 
was not present. The necessary conclusion is that the ears do perceive 
phase differences. 
The difference of phase effect ceased with increasing frequency at a 
frequency, of 1200 to 1500 d.v. This was tried with four observers. When 
the effect ceased the phantom sound remained in the median plane with 
changing phase. 
Intensity and phase combined. — Experiments in which phase and in- 
tensity differences were simultaneously applied showed that the actual 
relative importance of either of the two does not change in the combi- 
nation. Thus the effect of an intensity difference at the ears merely made 
a relatively slight change in the angular displacement of the phantom 
source. Consequently our conclusions formed by phase difference altera- 
