Vol. 8, 1922 
PSYCHOLOGY: D. MACKENZIE 
191 
Twenty observers (ten men and ten women) using each one ear made 
observations with the thermal receiver over a range of frequencies from 
100 to 4000 cycles per second, and in each comparison over a range of 
10 to 1 in acoustic pressure. Their results are expressible in a simple 
logarithmic law: log Pi = A + B log P2, where Pi and P2 are the alter- 
nating pressures of two tones phonometrically balanced. This result 
is concordant with the simple formulation of Fechner's law. 
The simple statement of Fechner's law holds only over the range of 
intensities throughout which the least perceptible stimulus increment is 
constant. Since the thermal receiver did not cover a wide range (the ear 
readily accomodates to acoustic pressures between 1/1000 and 1000 dynes 
per sq. cm.) the investigation was extended over wider limits of loudness, 
substituting an electromagnetic receiver as sound generator. 
Only four observers took part in this work. Some consideration was 
devoted to the types of curves to be expected if the ear is presumed to 
resemble the eye in its variation in response to varying values of the 
stimulus. The expectation was that two types of curves showing the 
relation between the alternating pressures of the two tones compared 
would be met with; sensibly straight throughout, and curved in a specific 
way. 
Tones of the following frequencies were compared directly with the 700 
cycle tone: 200, 300, 500, 1000, 1500, 2000, 3000 and 4000 cycles. At 
levels of loudness from 1 to 10 napiers above the threshold (one napier 
corresponds to a pressure ratio of ^ to 1) phonometric comparisons were 
satisfactorily made. The results were as anticipated: of 32 curves, 25 
were sensibly straight with slope unity over the whole range, 7 were curved 
oppositely in upper and lower halves. 
The acoustic output of the receiver when held against the ear cannot 
be directly measured. Systematic errors, therefore, affect the receiver 
calibration obtained under artificial conditions. Nevertheless, these 
errors are the same at all intensities, and the predominance of straight 
phonometric lines of unit slope is evidence for the conclusion that the rela- 
tive sensitivity of the ear is practically invariable over the range from near 
the threshold to near the level of painful loudness. 
The individual differences among observers are smaller at intermediate 
levels than at the threshold, and no significant difference is found between 
the ears of men and those of women. Fig. 2 exhibits the relative ear 
sensitivity derived from the observations of 4 observers using the thermal 
and the electromagnetic receivers. 
* A flicker phonometer similar in principle to the flicker photometer has been patented 
by C. M. Heck: U. S. Pat. No. 1,356,359; Oct. 19, 1920 "Method and Apparatus for 
for Comparing Sounds." 
