The rise of Sirius was carefully observed by the islanders of Ceos, in the Aegean. They watched for it with arms in their hands and sacrificed on the mountains to the star, drawing from its aspect omens of the salubrity or unhealthiness of the coming year. The sacrifice was believed to secure the advent of the cool North winds (the Etesian winds as the Greeks call them), which regularly begin to blow about this time of the year, and mitigate the oppressive heat of summer in the Aegean. See Apollonius Rhodius, _Argon._ ii. 516-527, with the notes of the Scholiast on vv. 498, 526; Theophrastus, _De ventis_, ii. 14; Clement of Alexandria, _Strom._ vi. 3. 29, p. 753, ed. Potter; Nonnus, _Dionys._ v. 269-279; Hyginus, _Astronomica_, ii. 4; Cicero, _De divinatione_, i. 57. 130; M. P. Nilsson, _Griechische Feste_ (Leipsic, 1906), pp. 6-8; C. Neumann und J. Partsch, _Physikalische Geographie von Griechenland_ (Breslau, 1885), pp. 96 _sqq._ On the top of Mount Pelion in Thessaly there was a sanctuary of Zeus, where sacrifices were offered at the rising of Sirius, in the height of the summer, by men of rank, who were chosen by the priest and wore fresh sheep-skins. See [Dicaearchus,] “Descriptio Graeciae,” _Geographi Graeci Minores_, ed. C. Müller, i. 107; _Historicorum Graecorum Fragmenta_, ed. C. Müller, ii. 262. 71 Above, pp. 24 _sq._ 72 We know from Censorinus (_De die natali_, xxi. 10) that the first of Thoth coincided with the heliacal rising of Sirius on July 20 (Julian calendar) in the year 139 A.D. Hence reckoning backwards by Sothic periods of 1460 solar years we may infer that Sirius rose on July 20th (Julian calendar) in the years 1321 B.C., 2781 B.C., and 4241 B.C.; and accordingly that the civil or vague Egyptian year of 365 days was instituted in one of these years. In favour of supposing that it was instituted either in 2781 B.C. or 4241 B.C., it may be said that in both these years the rising of Sirius nearly coincided with the summer solstice and the rising of the Nile; whereas in the year 1321 B.C. the summer solstice, and with it the rising of the Nile, fell nineteen days before the rising of Sirius and the first of Thoth. Now when we consider the close causal connexion which the Egyptians traced between the rising of Sirius and the rising of the Nile, it seems probable that they started the new calendar on the first of Thoth in a year in which the two natural phenomena coincided rather than in one in which they diverged from each other by nineteen days. Prof. Ed. Meyer decides in favour of the year 4241 B.C. as the date of the introduction of the Egyptian calendar on the ground that the calendar was already well known in the Old Kingdom. See L. Ideler, _op. cit._ i. 125