Along with A and B batteries, many older vacuum tube-based radios sometimes required the use of a C battery to provide a source of bias voltage for the grid of the tube or tubes employed in the radio. These tube types had no separate cathode element, and therefore could not be biased by the then-usual method of a cathode-bias resistor.
Unlike the voltages found on the other elements of the tube, the grid bias voltage was negative with respect to circuit ground or the B- voltage. It served two purposes: to set the operating characteristics of the tube, and to limit the amount of plate current flowing through the tube (which also helped lengthen the life of the B battery). C batteries were available in standard voltage ratings of –3, –4.5, –6, –7.5, and –9 volts and resembled ordinary batteries, though they were usually in rectangular form. Radio engineers could also use a device known as a "bias cell" – a small circular battery, resembling modern watch batteries and having a potential of one volt, to serve as a source of bias voltage. These bias cells were found not only in battery-operated radio sets, but also in some AC radios where a non-fluctuating source of bias was required.
In radio receiver use, a C battery could last indefinitely because it served only as a source of potential and not of current. Only deterioration of the battery’s characteristics could warrant its replacement, and in service they were usually only checked on a yearly basis.
By the end of the 1930s, more efficient tubes appeared on the market that required little or no grid bias voltage, and what might be required could be supplied by good circuit design. The C battery virtually vanished from radio receivers, much sooner than the A or B batteries.
Ghirardi, Alfred A., Modern Radio Servicing
. New York
: Radio & Technical Publishing Co., 1935.
Kendall, Lewis F, Jr., and Koehler, Robert Philip, Radio Simplified
: The John C. Winston Company, 1925.