Now the assumed value of the word translated year (viz. 12 hours), in its application to the passages just quoted, gives for the powers of the three terms three divisions of time as natural as could be expected under the circumstances.
1. [a]Σώσσος].—The sosus=30 days and 30 nights, or 12 hours × 60, or a month of 30 days, [a]μὴν τριακονθήμερος]. Aristotle writes—[a]ἡ μὴν Λακωνικὴ ἕκτον μέρος τοῦ ἐνιαυτοῦ, τοῦτο δέ ἐστιν ἡμέραι ἑξήκοντα].—From Scaliger, De Emendatione Temporum, p. 23. Other evidence occurs in the same page.
2. [a]Νῆρος].—The nerus=10 sosi or months=the old Roman year of that duration.
3. [a]Σάρος].—The sarus=6 neri or 60 months of 30 days each; that is, five proper years within 25 days. This would be a cycle or annus magnus.
All these divisions are probable. Against that of 12 hours no objection lies except its inconvenient shortness. The month of 30 days is pre-eminently natural. The year of 10 months was common in early times. In favour of the sarus of five years (or nearly so) there are two facts:—
1. It is the multiple of the sosus by 10, and of the nerus by 6.
2. It represents the period when the natural year of 12 months coincides for the first time with the artificial one of 10; since 60 months=6 years of 10 months and 5 of 12.
The historical application of these numbers is considered to lie beyond the pale of the present inquiry.
In Suidas we meet an application of the principle recognised by Rask, viz. the assumption of some period of which the sarus is a fraction. Such at least is the probable view of the following interpretation: [a]ΣΆΡΟΙ]—[a]μέτρον καὶ ἁριθμὸς παρὰ Χαλδαίοις, οἳ γὰρ ρκ´ σάροι ποιοῦσιν ἐνιαυτοὺς βσκβ´, οἳ γίγνονται ιε´ ἔνιαυτοὶ καὶ μῆνες ἕξ].—From Cory's Ancient Fragments[3].