Scientific American Monthly, October, 1920, page 178:

RADIO  MUSIC

    IT has been well known for some years that by placing a form of telephone transmitter in a concert hall or at any point where music is being played the sound may be carried over telephone wires to an ordinary telephone receiver at a distant point, thus enabling those several miles away to listen to the music. Such systems have been in use in London between a number of the theaters and hotels for many years, but it is only recently that a method of transmitting music by radio has been found possible.
    It has now been discovered that music can be transmitted by wireless in the same manner as speech or code signals and as a result of research work on radio telephony at the Bureau of Standards it has been proven that music sent by this means does not lose its quality. It is, therefore, obvious that music can be performed at any place, radiated into the air through an ordinary radio transmitting set and received at any other place, even though hundreds of miles away. The music received can be made as loud as desired by suitable operation of the receiving apparatus. The result is perhaps not so very different from that secured by means of the ordinary telephone apparatus above mentioned, but the system is far simpler and does not require the use of any intermediate circuit. The entire feasibility of centralized concerts has been demonstrated and in fact such concerts are now being sent out by a number of persons and institutions. Experimental concerts are at present being conducted every Friday evening from 8:30 to 11:00 by the Radio Laboratory of the Bureau of Standards. The wave length used is 500 meters. This music can be heard by any one in the territory near the District of Columbia having a simple amateur receiving outfit. The possibilities of such centralized radio concerts are great and extremely interesting. One simple means of producing music for radio transmission is to play a phonograph into the radio transmitter. An interesting improvement upon this method is being utilized in the experiments at the Bureau. The carbon microphone, which is the mouthpiece of an ordinary telephone, is mounted on the phonograph in place of the usual vibrating diaphragm. As a result the phonograph record produces direct variations of electric current in the telephone apparatus instead of producing sound; thus while the music is not audible at the place where the phonograph record is being played, it is distinctly heard at the different receiving stations.