440 SWAN. 



normal conditions may be roughly estimated to be at least twenty- 

 five times the average intensity which would exist with totally absorb- 

 ent walls. Hence, as the current in ^Yien's telephone was reduced 

 until the sound was just audible at a distance of thirty centimeters, 

 only a fraction reaching the ear came in a direct path from the tele- 

 phone, the balance being contributed by the reverberation in the room. 

 In other words, at the limit of audibility the current in the telephone 

 was less than would have been required for minimum audibility in a 

 non-reverberant space, thus giving an apparent sensitiveness greater 

 than actually was the case. 



Furthermore, with a fixed relation between the source and the ear, 

 as in Wien's experiments, the interference system in the room is not to 

 be disregarded, as the ear cannot conceivably be at a point of average 

 disturbance for all pitches with a constant distance between ear and 

 telephone, and is not likely to be at such a point for even a single pitch. 



Add to this the reaction of the room upon the source, making it 

 impossible to determine the amount of energy emitted by the telephone 

 diaphragm or even to compare sources of different pitch, and the 

 wonder is, not that Wien's results are too small, but that there is any 

 possible coincidence that should have caused them to fall on anything 

 approaching a regular curve. 



The claim has been made that the wide discrepancies in previous 

 investigations are due to inadequate apparatus and to differences in 

 hearing in the normal ear. These criticisms are not altogether justi- 

 fied, however, and the true causes of the variations are undoubtedly 

 due to the reasons already discussed. Differences between ears, even 

 so-called normal ears, do exist in some cases, no doubt. Even a 

 trained ear may vary from night to night owing to nervous fatigue or a 

 slight cold. But experience for many years with the average results 

 of a number of trained observers in measuring reverberation has shown 

 a surprising conformity in threshold sensation. One cannot help but 

 wonder if the differences noted by some experimenters is not due 

 largely to unskilled observation. The sense of hearing must be trained 

 just as the sense of taste in the case of a tea taster. It is assuredly 

 impossible to call upon inexperienced students or members of the staff 

 to make a few casual observations with any hope of consistency. 



The values of the minimum audible intensity obtained by Fletcher 

 and Wegel are exceedingly interesting, as they are free from most of 

 the sources of error we have noted. They are obtained, however, from 

 the averages of forty-one persons, both men and women, many, if not 

 most of whom, could hardly have been experienced in detecting the 



