One evening in the winter of 1873-4 I came home from my laboratory work and went into the bathroom to make my toilet for dinner. I found my nephew, Mr. Charles S. Sheppard, together with some of his playmates, taking electrical "shocks" from a little medical induction-coil that I heard humming in the closet. He had one terminal of the coil connected to the zinc lining of the bathtub--which was dry at that time--while he held the other in his left hand, and with his right was taking shocks from the lining of the tub by rubbing his hand against the zinc. I noticed that each time he made contact with the tub, as he rubbed it for a short distance, a peculiar sound was emitted from under his hand, not unlike the sound made by the electrotome that was vibrating in the closet. My interest was immediately aroused, and I took the electrode out of his hand and for some time experimented with it, going to the cupboard from time to time to change the rate of vibration of the electrotome, and thus change the quality of the sound. I noticed that the sound or tone under my hand, if it could be so called, changed with each change of the rate of vibration. The thing that most interested me was that the peculiar characteristics of the noise were reproduced. In those few minutes I laid out work enough for years of experiment, and as a result I was late to dinner. This discovery opened up to my mind the possibility of three things--the transmission of music and of speech or articulate words through a telegraph-wire, and the transmission of a number of messages over a single wire. I constructed a keyboard consisting of one octave and made a set of reeds tuned to the notes of the scale, and then when some one would play a melody I could reproduce it in two ways: One by placing my body in the circuit and rubbing a metal plate--it might be the bottom of a tin pan, a joint of stovepipe or otherwise--anything that was metal and would vibrate would give the effect. Another way was to connect an electromagnet (having a diaphragm or reed across its poles) in the circuit at the receiving-end and mount it on some kind of a soundboard. I made a great number of different kinds of receivers that were capable of receiving either musical or articulate sounds, as has many times been proven by experiment. I carried two sets of experiments along together; the one looking toward a system of multiple telegraphy and the other the transmission of articulate speech. Let us first look into the multiple telegraph and take the other up under the head of the telephone.