The signs used for the tetramenes are supposed to represent water, a field with growing plants, and a barn; the natural order would be that the first should represent the inundation, the second the sowing which succeeds it, and the last harvest-time. If this be conceded, the initial system would have had the month Thoth connected with the water sign, as Thoth in early Egyptian times was the first inundation month. But in the times of the Ramessids even this is not so. Thoth has the sowing sign assigned to it. In the time of the Ptolemies the Hood is no longer in Thoth, but in Pachons, and Pachons has the barn sign attached to it, while the month Thoth is marked by the water sign, thereby bringing back the hypothetical relation _between the name of the month and the sign_, although, as we have seen, Thoth is no longer the flood month. Egyptologists declare that all, or at least part, of this change took place between the periods named; they are undoubtedly justified as regards a part. At one point in this interval we are fortunately supplied with some precise information. In the year 238 B.C. a famous decree was published, variously called the decree of Canopus and the decree of Tanis, since it was inscribed on a stone found there. It is perfectly clear that one of the functions of this decree was to change, or to approve an already made change in, the designation of the season or tetramene in which the inundation commenced, from Thoth to Pachons. Another function was to establish a fixed year, as we shall see presently. We must assume, then, that a vague year was in vogue prior to the decree. Now the decree tells us that at its date the heliacal rising of Sirius took place on 1 Payni. Assuming that this date had any relation to the system we have been considering, the cycle to which it belonged must have begun Days. 5 Epacts 30 Mesori 30 Epiphi 30 Payni ── 95 × 4 = 380 years previously--that is, in the year 618 B.C. Here at first sight it would seem that the Sothic cycles we have been considering have no relation to the one now in question; for, according to my view, the last Sothic cycle began in 1728 B.C. A little consideration, however, will lead to the contrary view, and show that the time about 600 B.C. was very convenient for a revision of the calendar. In the first place, nearly a month now elapsed between the coming of the flood and the heliacal rising; and in the second, by making the year for the future _to begin with the flood_, a change might be made involving tetramenes only. Thus, commencement of cycle 1728 B.C. Epacts 5 Two tetramenes 240 Month between flood and rising of Sirius 30[78] ——— 275 × 4 = 1100 ———— 628 B.C. Nor is this all. A very simple diagrammatic statement will