That marvellous conception had slowly flowered through a synthesis of ideas and observations made in the course of his work on the harmonic telegraph. After he had given up on the tuning forks and had started to think in terms of organ reeds, he began to contemplate larger numbers of transmitting units. He knew from his musical experience that if he put his head inside a piano and sang or spoke, a number of strings would respond. Hence if he constructed a “harp transmitter” with enough strings or reeds to pick up every frequency of the human voice, their combined vibrations could be converted into a complex electric current that would vary in intensity with the varying sounds of the voice. And a receiver harp at the other end of the circuit would reproduce those sounds. Although Bell suspected that his theoretical harp transmitter was too big, too complicated, and probably too expensive to be practical, he nevertheless felt that the underlying principles were sound, and the conception persisted in his mind.

Meanwhile, his work with the deaf had taken him one day to the physics laboratory at the Massachusetts Institute of Technology to inspect a remarkable new instrument called the phonautograph. This contrivance was a kind of speaking trumpet, closed at the far end by a stretched membrane. Attached to the membrane was a stylus. When words were spoken into the mouthpiece the membrane vibrated, causing the stylus to trace an oscillating wave pattern on a piece of smoked glass. Bell thought that the instrument might be useful in teaching articulation to the deaf by revealing to them visually the relationship between the sounds they articulated, or tried to articulate, and the patterns traced by the stylus on the smoked glass.

Unfortunately the phonautograph did not work satisfactorily for Bell’s purposes. He was struck, however, by a similarity between its mechanism and that of the human ear, and it occurred to him that the phonautograph might be improved were it modelled more closely upon the structure of the ear. Seeking more accurate anatomical information, he consulted a famous Boston ear specialist, Dr. Clarence J. Blake. Somewhat to Bell’s surprise, Blake suggested that instead of constructing a phonautograph around a model of the ear, he should use an actual human ear, excised from some donor in the morgue. What was more, the Doctor would provide one. And he did, properly preserved and prepared for scientific study. The experiment proved highly successful. Bell constructed a new phonautograph, using the ear as a component, and found that its tracings of sound patterns on the smoked glass were vastly more accurate than those of the instrument at M.I.T. But what stirred Bell more deeply than this hope of a new device for teaching the deaf was the opening of a new avenue of thought provided by his glimpse into the secret corridors of the inner ear.

Bell carried his apparatus home with him to Brantford in July, 1874, and continued his experiments with it during his summer vacation. He continued to marvel over the mechanisms of the ear, and especially the ability of the tiny diaphragm—the ear drum—to move the relatively heavy bones of the ear. Then suddenly, on July 26, 1874, one of those amazing cross-circuits of thought that happen without premeditation produced a blinding scintillation in his mind. For several hours he had been brooding over the problems inherent in his harp transmitter, wondering if he could not find a less cumbersome device than one involving a whole choir of strings or reeds—and some simpler way to pick up the sounds of the human voice and generate a current that would vary in intensity as the air varied in density during the production of those sounds.

Years later Bell described the exact moment at which he suddenly perceived the solution he had sought: “I do not think that the membrane of the ear could have been half an inch in diameter and it appeared to be as thin as tissue paper. . . . It occurred to me that if such a thin and delicate membrane could move bones that were, relatively to it, very massive indeed, why should not a larger and stouter membrane be able to move a piece of steel in the manner I desired?”

Bell knew now he could discard all the capricious and multitudinous reeds in his harp transmitter. A single diaphragm could take their place. And a single magnetized reed, attached to the center of the diaphragm and vibrating with the sound of the human voice, could generate a current that would vary in intensity precisely as the air varied in density during the production of that sound.

“At once the conception of a membrane speaking telephone became complete in my mind,” Bell related, “for I saw that a similar instrument to that used as a transmitter could also be employed as a receiver.”