We notice also that the bird over the first tree, although differing in some respects from it, is the same as that in the top or red loop of the other plate, and that over the third tree the same as that in the blue or bottom loop, agreeing also in this respect. From these facts we understand that the black figure is sometimes at least assigned to the south. I am fully aware of the difficulties to be met with in attempting to carry out this assignment of colors, in explanation of other plates of this and other Codices, nor do I believe colors can be relied upon. They form some aid in the few plates of general application to the calendar, and where there are reasons, as in the cases given, to suppose the cardinal points will be indicated in some regular order. The same thing is true also in regard to the Manuscript Troano. For example, if we suppose character _a_ of Fig. 7 to denote the east, _b_ north, _c_ west, and _d_ south, we shall find them arranged in the following different ways: ______ ______ | | abcd cdab | | | c b | | c a | | | | | | d a | | d b | |______| |______| ______ ______ | | | | | a d | | c d | | | | | | c b | | a b | |______| |______| Combine with these colors and other distinctive marks, then vary them in proportion, and we should have an endless variety, just as we see in the Mexican Codices. We can only hope to solve the problem, therefore, by selecting, after careful study, those plates which appear to have the symbols arranged in their normal order. Turning to plate 43 of the Borgian Codex, we find it impossible to make it agree, either with the plate of the Fejervary Codex or the Vatican Codex. Here we find the days 1, 5, 9, 13, 17 associated with the green figure in the lower left-hand square; 2, 6, 10, 14, 18 with the yellow figure in the lower right-hand square; 3, 7, 11, 15, and 19 with the black figure in the upper right-hand square, and 4, 8, 12, 16, 20 with the red figure in the upper left-hand square. What adds to the difficulty is the fact that the symbol of the _Cane_ accompanies the black figure, thus apparently indicating that this denotes the year Acatl. That these groups are to be taken in the same order as those of Plate 44 of the Fejervary Codex, that is around to the left, opposite the sun's course, is evident from the days and also from Plate 9 of this (Borgian) Codex, where the twenty days of the month are placed in a circle. In this latter the order of the four years is indicated by the first days of the years, viz, _Cipactli_, _Miquiztli_, _Ozomatli_, and _Cozcaquauhtli_ placed in blue circles at the corners in the following order: