_Variation of major units:_ Was variation more allowable in the major than in the minor unit of a series? This was tested first in the same manner as for accents. In a series of which a double string was the major element, a third string was hung with every such double, thus changing the unit in both size and content. [Illustration: Fig. 8] The series immediately readjusted itself with the three-group as element for most of the subjects, although one was still able to feel them all as one unit, varied by the added string. Varying only the _size_ of the major unit gave the same result. [Illustration: Fig. 9] The pairs, instead of remaining the same size, were made alternately larger and smaller, and a new repetition was made, _i. e._, with the larger pair as major element and the smaller one as minor. They all agreed, however, that less change was made in these cases than when the _minor_ spaces had been changed in size. In the latter case either a regrouping was made, changing the whole character of the series, or it was spoiled altogether. With change of _major_ units alone, however, although a new element was made, it was still possible to take it in the old way without much difficulty or change in feeling-tone. It was then necessary to see how change of content would affect the major unit, the size remaining constant. A group of two sets of double strings 10 ccm. apart was taken as the repeated element, and these groups placed at 10 ccm. from each other. [Illustration: Fig. 10] Within one element was hung one string, and within the next two, thus varying the content while the size remained constant. In every case the answer was the same. It was not so pleasant as when the filling was the same, but the group still remained the unit of the repetition, and the series essentially the same. Several variations were made in this element. Instead of hanging strings regularly (1 in one, and 2 in the next) they were hung irregularly, _i. e._, an extra one here and there at intervals in no special order. As long as the boundary-lines of each group remained at the same distance from each other, and from the next group, thus keeping the unit at uniform size, although the pleasure-tone varied, the balance of the series was not changed. No regrouping or shifting of the equilibrium resulted. It would seem from the preceding experiments that in any series variation of the _major unit_ was tolerated more than of the _alternate;_ while in _either case_ variation in _content_ had less influence than variation in _size_.