THE interesting question raised by Prof. F. Allen's letter in Nature of July 21 is whether the difference tones he observes are present in the air or whether they exist only in the auditory system of the listener. If the loudspeaker to which the two oscillators are connected (whether in series or parallel is only a matter of impedance matching) is strictly linear in its amplitude response, then no combination tones, sum or difference, will be produced. The ear, on the other hand, is far from linear in its amplitude response; so when two high-pitched tones are heard together, the difference tone is often clearly audible while the sum tone may be less easily distinguished or above the limit of audibility, depending on the pitch of the beating tones. However, if the loudspeaker is nonlinear, combination tones will be physically present in the air nearby and could be detected by a linear tuned sound analyser. If this is not available it should be possible to feel or even see the low frequency vibration of the loudspeaker diaphragm when the two tones are nearly in synchronism.