While carrying on a study of intonation in ensemble string performance, it was discovered that the degree of amplitude modulation present in tones played with frequency modulation (vibrato) varied greatly from tone to tone in a melodic line played at a fairly constant level of intensity. A study of a large number of tones played with vibrato showed that while the rate of amplitude modulation for any given tone coincided with rate of frequency modulation for that tone, there was no such coincidence in range. It was found that the peak of intensity of a cycle of amplitude modulation could occur, with different pitches, at any point along the range of a cycle of frequency modulation. This seemed to indicate that these amplitude modulations were caused by the varying characteristics of frequency response of the string instruments involved. A series of tests made in an anechoic chamber involving cellos, revealed that this was indeed the case and that the amplitude modulations resulting from response characteristics are a very important factor in distinguishing the vibrato from the nonvibrato tone.