The apparatus consists of a couple of glass tubes with electrically-heated wire filaments about one quarter of the way up the tubes. It is thus related to the Chemical Harmonica. As the heated air rises, cooler air is drawn up into the bottom, resulting in the formation of eddies in the air. A wide range of frequencies is thus present, and the frequency whose half wavelength is equal to the length of the tube (plus a small end correction) will resonate. 

   The shorter of the tubes has a paper sleeve placed on its upper end. This may be slid up and down, causing the frequency to rise and fall. In demonstrating beats, I start with the sleeve all the way down. The frequency is thus somewhat higher than that produced by the other tube, and beats are produced. As the sleeve is slide up the tube, the frequency goes down, and eventually the zero beat condition is obtained. Sliding it still further up produces beats once more, showing that the important quantity is the absolute value of the difference between the two frequencies. 

   The apparatus was in the Central Scientific Company catalogue for many years. In 1940 it cost $31.00.