The Microphonic Principle
Although the underlying principle is clear now, it was not fully understood when speaking and listening with the first telephones was more art and imagination than science - mainly due to poor transmission (Harlow 1936)
Regardless of who actually invented the microphonic transmitter, "The principle of the operation of microphonic contacts was made known to the scientific world by Prof. David E. Hughes in England, in May 1878" (Rhodes 1929). What Hughes described, and what all other microphonic experimenters worked on, was something "electricians have hitherto scrupulously avoided" - a bad contact (Wiles 1926) or "the microphonic nature of loose contacts between metals and other conducting substances" (Rhodes 1929)
"Hughes, while experimenting to test something else, broke a piece of stretched wire connected to a telephone receiver.. and it was only when the wire vibrated so strongly as to break , that he heard a sound at the moment of its fracture. When he again joined the two ends of wire, another sound was produced, and he soon perceived that imperfect contact between the two broken ends of the wire would enable him to obtain a sound. " (DuMonciel, 1879). Hughes talked to all sorts of conducting items, in loose contact on a hard surface. Piles of watch chain, nails crossing one another, and pieces of carbon all worked! The results in an ordinary "Bell" electromagnetic receiver were astonishing.
The vital point is that the contacts must touch at all times, regardless of pressure changes - breaking contact causes rattling in the receiver and stops proper speech transmission, because of the complex nature of voice sounds - amplitude and timbre cannot be transmitted by contacts only opening and shutting with the frequency of voice waves (like a Morse key).
Carbon was discovered to be the best practical material - a conclusion reached also by Edison, but for different reasons. In one celebrated test, the movement of a fly walking on the sounding board caused sufficient vibrations of loosely-contacting carbon pieces to make clear sounds in the receiver (see Fig 7) being Hughes' classic, simplest model. Carbon heated up in an atmosphere of vaporized mercury worked even better. DuMonciel (1879) wrote "Mr Hughes has tried other substances, including those which are good conductors, such as metals. Iron afforded rather good results, and platinum.. was equal to , if not superior to, mercurised carbon he prefers the carbon, which (is) incapable of oxidation".
It was this realization that ANY two conductors in loose contact would transmit which showed Hughes correctly understood the principle - slight movements in response to voice waves created tiny forces, minutely varying the amount of contacting surface area . This changed the resistance of the contact(s) , varying current flow out of all proportion to the physical changes. But this principle, used by Edison and explained by Hughes, "was (first) discovered in 1856 by Count du Moncel (who stated) that pressure exercised at the point of contact between two conductors touching one another had a considerable influence on the (current flow) . " It was successfully applied by Clerac, a French Telegraph officer. (Hosrpitialer, 1883)
Hughes was however wrong to call the device a "microphone" (he coined the term) , thinking it actually amplified sounds. Later, Berliner and Edison both put the varying microphone current through the low-resistance primary winding of a step-up transformer (coil) which did increase output. (The single-contact types, particularly, would burn the points or fuse them if too much current was put directly through them.)
Edison's work was along the same lines (see his commercial form of transmitter, as cased by G M Phelps, see Fig. 6, but he started with carbon and similar substances, called by him "semiconductors". This term alone shows that he incompletely understood the nature of the principle. He thought that compression and release of powdered carbon or similar substances, in response to voice pressure, varied resistance.
Nonetheless, his transmitter using a contact of platinum against carbon worked fairly well for the time - much better than the feeble electromagnetic Bell instrument. As such it (and Edison with it) was snapped up by the giant Western Union, who set out to steamroll the struggling Bell interests and corner the new telephone business, before being brought to heel by the courts. Eventually Western Electric was spawned out of the settlement deal, manufacturing most of Bell telephone's equipment from 1881 onwards.
This German inventor remains controversial. Fig 1 is taken from Thompson (1883) - still the best reference on Reis. Numbers 1 to 5 show schematics of Reis' main models (Thompson identifies 9 in total). Although all differently constructed, each had the all-important junction. Reis tried various metals - never carbon, which some say was why he never made one "talk" for too long, and then not too well. (Prescott, 1884). Thompson, scorning Edison's "semiconductor" carbon theories, heard better articulation through pure silver contacts than any carbon junction - which he says fits Hughes' (correct) theory; and also shows that Reis was completely on track. (Thompson, 1883 and also Aitken, 1939).
Number 4 in the figure shows Reis's contact adjusting screw, also in 3. Thompson says this proves Reis knew the importance of a varying contact that did not break and make - which is the objection always raised, because Reis' documentation was poor, and translations all quote him as saying to adjust for making and breaking as the voice waves move the diaphragm. Even his final form (No. 5) which operated horizontally without the screw allows varying contact by combined inertia of the upper contact piece and gravity. Compare that with Berliner's (vertical) model in No. 6. (see Fig 4) which definitely relied on a heavy hinged spring and carbon block to stop contact-breaking. Reis' No. 3 and 5 were actually tested (with stepup transformers and Bell receivers) by the British Post Office in 1932, and were both found to transmit speech, though poorly at low volume. Adjustment was critical and kept going out - but speech was sent. (Full report in P.O.E.E. J. vol 25, 1932) . This backs a lot of period reports of speech transmission by Reis.
Note also the similarities between Reis models and the Blake, No. 7. The last 3 numbers are all Edison models - 9 being the production carbon buttontype - see Fig 6. In all types the resemblance is clear - despite varying construction details and materials, or even use of double contacts. Reis does seem to have missed something vital though - he was best known for transmitting accurate musical tones , and "the often successful transmission of words
was due to accidental adjustment of his contacts to a true microphonic condition" (Aitken 1939).