82 PROCEEDINGS OF THE AMERICAN ACADEMY 



brane of paper carrying a small electro-magnet, with its poles facing 

 the poles of the large inducing magnet. As this has its polarity 

 determined by induction, and the strength of it varying with the dis- 

 tance apart of these facing poles, it is evident that any articulate re- 

 sults obtained under these conditions must be due to the ordinary 

 vibrations to and fro of the disk and magnet upon it. The results 

 with this arrangement were excellent ; the articulation was not only 

 good, but it was as loud as in any telephone now in the market. 

 Especially good was it when the electro-magnet was separated from 

 the large magnet by a rather thin piece of rubber, allowing of motion 

 in a longitudinal direction, but not in any other. This I take to be 

 demonstrative evidence that the explanation given above of the tele- 

 phonic action is true. Diaphragms of other materials were tried, 

 such as wood, zinc, brass, mica, and so on, — a great number. All will 

 work ; and paper was found to give the best results, probably because 

 it has rigidity and elasticity enough, while the mass is small, and hence 

 more easily moved by aerial vibrations. 



At one time, early in October, it occurred to me to try for a receiver 

 the iron pole of one of the permanent magnets, with its bobbin on of 

 course. That is to say, the permanent magnet was removed, the 

 other conditions remaining as before. With this I was able both to 

 send and receive. Then the iron core was removed, and in its place 

 an ordinary wood-screw an inch and a quarter long was put through 

 the bobbin and screwed into a piece of board three or four inches 

 square, the head of the screw towards the plate. With this arrange- 

 ment I was still able to receive or send, using the circuit between the 

 College and my house, nearly half a mile long. Now every wood- 

 screw is made slightly magnetic by the process of making the screw, 

 but most wood-screws have but a slight polarity ; that used in this 

 experiment was only sufficient to move slightly three or four iron 

 filing bits. Again, experiments were tried with magnets of various 

 forms, — magnets with chambered poles, with poles cut radially, with 

 poles bent at right angles, with wire of various sizes, from No. 22 to 

 No. 37, with the result that No. 28 or 30 seemed preferable for nearly 

 every purpose. The finer wire not only offers so much greater 

 resistance, but in the summer-time it is liable to be injured by light- 

 ning. I have had several fine- wire telephones that were injured by 

 the induced current from a flash of lightning a considerable distance 

 away. My line runs for thirty or forty rods under the Somerville 

 fire-alarm wires, and hence would be affected by the inductive action 

 upon them ; but between the college building and my house, there is 



