One way to make a carbon microphone is to rest a piece of pencil 'lead' (that is graphite + a binding material) a few centimetres long on the edges of two razor blades. The set-up is [itex]\pi[/itex] shaped, with the razor blades forming the uprights, and the pencil 'lead' forming the bridge across the top. The 'terminals' of the microphone are the razor blades.
Connect the microphone in series with a battery (6 V, perhaps) and a resistor (100 [itex]\Omega[/itex], perhaps. Connect the y-amplifier input terminals of an oscilloscope across the resistor. Speak loudly near the pencil lead and you should see a corresponding trace on the oscilloscope, when you've adjusted the time-base and the y-gain.
It works - if it does work! - by the pressure variations in air due to the sound waves pushing the 'lead' more and less firmly on to the blades, changing the contact resistance between blades and lead. The series resistor completes a potential divider circuit.
Until quite recently carbon microphones (using carbon granules) were the standard microphone in telephone handsets. The quality of reproduction was never good.