Further investigation of primitive music requires a study of the intervals themselves. Notwithstanding the great differences of systems we find that all intervals may be interpreted as subdivisions of the octave. To the untrained ear the octave appears very commonly as a single tone; in other words, no distinction is made between a tone and its octave. To a lesser extent this is true of the fifth and even of the fourth. The majority of intervals that have been found must be considered as subdivisions of the octave. However, the subdivision does not always proceed according to harmonic principles as in our music, but by equidistant tones. The development of harmony in modern music has had the effect that we have lost all feeling for equidistance in a harmonic series and that the recent music in which non-harmonic equidistant tones are applied require a difficult break with the pattern of musical form to which we are accustomed. After a long struggle, we have reached a compromise between the two systems, the harmonic and the equidistant, by dividing the octave into twelve equal parts which give a fairly close agreement to the natural harmonic intervals, although the differences are audible to a trained ear. The Javanese divide the octave into seven equidistant steps, the Siamese into five, systems that are in fundamental conflict with those of our music. In short, a great variety of scales exist and serve as foundation for the musical systems of different people. All seem to have in common as foundation the octave. I will not enter into this intricate subject any further, because a safe method has not yet been found that would enable us to tell definitely what people _want_ to sing among whom there is no theory of music, as it exists among ourselves or the civilized people of Asia, and who have no exactly constructed instruments.