Those who preferred this gave essentially the same reason. They agreed that the unity of both elements was broken up by this change, and they did not stand out distinctly from each other; but all felt a certain _congruity_ in having both major and minor units follow the same scheme in composition. They were not distinctly an alternating series, but harmonized better as lines. The two spatial subjects, who disliked this arrangement more than the other, gave the same reason: the unity of the elements was spoiled, they did not "hang together." Their dislike was similar in kind to that of the others, only the congruity which made up for it with the former failed to satisfy these. With the symmetry broken and the balance tipping in different ways, the feeling was not strong in either direction. They still criticised it in the same terms of congruity and distinctness, with no especial change on account of this modification. [Illustration: Fig. 18 A] [Illustration: Fig. 18 B] These experiments all pointed to the fact that (1) a certain amount of congruity and equality was necessary between elements of a series (although it did not establish what were the essential features of such a harmony). (2) It is more pleasant, as a rule, to have the elements symmetrical, although symmetry was not a necessity for an agreeable series. (3) Provided the change in the symmetry of the units was not enough to shift the whole order of the series, changing the major to minor units (and _vice versa_), any varying of the symmetry of a minor unit was more disturbing to the repetition than of the major, while varying their symmetrical position, as regards the unit on either side, was absolutely destructive to the order. The next experiment dealt with a different side of the question. Since the unit of a repeated series may be a group with repetitions inside itself, does the repetition of lines or figures _inside_ the group differ from the repetition of the groups as a whole? If so, how? That is, in the enjoyment of a series of groups with repeated lines in the group, in what respect does our apperception of the repetition differ in the two cases? Or does it in reality differ at all? To test this, the strings were arranged in the following way. 10 groups of five strings were hung 100 mm. wide and 100 mm. apart. Each unit had, then, five repetitions within it. [Illustration: Fig. 19] The arrangement was pleasant to all the subjects, and they described the effect of the experience, falling at once into the spatial and temporal types as before (this was wholly naïve, for the same questions were asked of each, and they had no idea of being grouped in types). The introspection of both types must be taken in some detail, to fully analyze the experience.