OF ARTS AND SCIENCES. 225 



spring rested on the vulcanite strip, in which case the pitch of the 

 fork was perfectly clear. A key was inserted so that the sound 

 produced at one or more successive contacts could be caused to 

 enter the ear at will. 



In our first experiments we employed as a source of sound a large 

 tuning-fork vibrating in front of a resonator and sounded by a ham- 

 mer. But we found that this gave too feeble a sound for our pur- 

 pose. We therefore made use of a small fork, which was set into 

 vibration by a blow, and whose stem was lightly pressed against 

 the disk of the transmitting telephone. A bit of wax placed at 

 the end of the stem was found to be sufficient to prevent any 

 chattering. 



The particular form of magneto-telephone which was found most 

 satisfactory as a transmitter, on account of its power, was a bipolar 

 instrument having a large diaphragm, a special form made by tbe 

 American Bell Telephone Company. For a receiver we found a 

 very sensitive form of magneto-telephone made in Sweden best 

 suited for our purpose. 



The tuning-forks employed were of small size, belonging to a 

 Valentine and Carr tonometer. In all of the experiments cited 

 the transmitter was placed in a distant room, so that no sound 

 could possibly reach the ear through the air. 



The speed of revolution of the brass wheel was read by a speed- 

 counter, and from this it was easy to determine the period of time 

 during which the contact-spring rested on the vulcanite sector. 



The mode of observation adopted consisted in driving the wheel 

 at a rate such that the difference in pitch between the brief sounds 

 heard in the receiving telephone seemed to be distinct. One of the 

 experimenters then sounded the forks successively, presenting them 

 one after another to the telephone, and noting the order in which 

 they were sounded. The listener at the receiver at the other end 

 of the line noted the pitch heard, and these estimations were after- 

 wards compared with the order as noted by the person at the trans- 

 mitter. The speed was then altered, generally being increased, 

 and a new set of experiments undertaken. In some cases, owing to 

 bad transmission or for other causes, the listener felt uncertain 

 whether the decision noted was correct, and in such cases the result 

 was marked as doubtful. 



In the first series of experiments only two notes were compared, 

 which were an octave apart; one, C 3 , of 256 vibrations; the other, 

 C 4 , of 512 vibrations. 



vol. xxvu. (n. s. xix.) 15 



