OF ARTS AND SCIENCES. 227 



15 33 67 



The following results were obtained with intervals other than 

 the octave. 



In the sets of experiments Nos. 22, 23, the contact-spring failed 

 to give even pressure ; and it was noticed that the sound was very 

 badly transmitted at certain times, which corresponded with those 

 at which some of the erroneous estimates were made. 



