ndicated by the first
days of the years, viz, _Cipactli_, _Miquiztli_, _Ozomatli_, and
_Cozcaquauhtli_ placed in blue circles at the corners in the following
order:

--------------------------------
|Ozomatli. Miquiztli.|
| |
|Cozcaquauhtli. Cipactli. |
--------------------------------

In the lower right-hand corner of Plate 4, same Codex, is a square with
the four quadrants very distinctly colored and arranged thus:

-------------------
|Yellow. Green.|
| |
| Blue. Red. |
-------------------

and a large red circle in the center, on the body of what is evidently
intended as a symbol of _Cipactli_. As this appears to be a figure of
general application, we presume that it commences with _Cipactli_, the
day on which the cycles began. As the four names of the days with which
the years began probably show, as arranged in the above square, their
respective positions in the calendar wheel, I infer that, in their
normal arrangement, _Cipactli_ corresponded with the red, _Miquiztli_
with the green, _Ozomatli_ with the yellow, and _Cozcaquauhtli_ with the
blue. This brings the colors in precise accordance with those on the
cross in the lower right-hand square of Plate 43; and if we suppose the
black figure to correspond with the blue it brings the colors in the
same order, but the day groups are shifted around one point to the left.
It is probable therefore that this plate, like a number of others in the
same Codex, is intended to denote the relation of colors and day groups
to each other in some other than the first or normal year, or possibly
to the seasons or the four Indications of the cycle.

But be this as it may, I do not think the difficulty in reconciling the
arrangement of the colors and days in this Codex will warrant the
rejection of our explanation of the plates of the other codices. That
Plate 44 of the Fejervary Codex is one of general application must be
admitted, as is also the "Table of the Bacabs" from the Cortesian Codex;
and if the true assignment to the cardinal points is made anywhere it
will certainly be in these. Turning now to the latter, as shown in our
Plate II, where the erased characters are restored, we note the
following facts, and then with some general remarks conclude our paper,
as we have no intention of entering upon a general discussion of the
Mexican Calendar, which would be ne