228 PROCEEDINGS OF THE AMERICAN ACADEMY 



It appears from the results given in the preceding tahles that 

 even with as small a fraction as -j* 2 q of a vibration of the lower fork 

 and y^o of a vibration of the higher fork, when these had an interval 

 of an octave, it was possible to distinguish one note from the other; 

 and when the duration of the tone was greater, but still much less 

 than two complete vibrations of the higher note, the distinction 

 became easy. Nor was the distinction possible merely because of 

 the possession of a clearly defined pitch by the higher note alone, 

 whose period was shorter, so that a greater number of vibrations 

 entered the ear in the time during which the circuit was completed 

 through the telephone; which, it might be surmised, would enable 

 one to distinguish it from the lower note even in the absence of any 

 clearly defined pitch for this latter. On the contrary, the ear al- 

 ways recognized the existence of a distinct pitch with each note, 

 even with the shortest duration of tone which was employed in the 

 experiments; but, as would be expected, the ability to determine 

 with certainty which of the two notes heard successively was the 

 higher in pitch diminished as the duration of the sound diminished, 

 and also as the interval between the two notes became less. The 

 results given in No. 24, Table VI., are curious, and at the time 

 they were reached seemed to be due to a misjudgment as to which 

 of the tones was the higher at the beginning of the experiment. 

 Series No. 25 was made immediately afterwards ; but the notes of 

 the forks were carefully listened to separately, in order, if possible, 

 to fix the pitch of each in the mind. As a result three fourths of 

 the estimates were correct. 



In another series of observations several forks were struck ir- 

 regularly, and an attempt was made to state which fork of the set 

 used was heard. The listener knew what were the forks compos- 

 ing the set. 



With forks C 3 , E 3 , G 3 , C 4 , the results given in Table VII. were 

 reached. The separate series of observations are separated by hori- 

 zontal lines. The table gives the name and rate of the forks used 

 in the first two columns. In column V is given the number of 



