Page:The history of Rome. Translated with the author's sanction and additions.djvu/237

Rh by days, which, as already mentioned, were counted not forward from the phase that had last occurred, but backward from that which was next expected; by lunar weeks, which id in length between 7 days and 8, the average length being 7$3⁄8$; and by lunar months, which in like manner were sometimes of 29, sometimes of 30 days, the average duration of the synodical month being 29 days 12 hours 44 minutes. For some time the day continued to be with the Italians the smallest, and the month the largest, division of time. It was not until later that they began to distribute day and night respectively into four portions, and much later still when they began to employ a division into hours; which explains why even stocks otherwise closely related differed in their mode of fixing the commencement of day, the Romans placing it at midnight, the Sabellians and the Etruscans at noon. No calendar, of the year at any rate, had as yet been organized when the Greeks separated from the Italians, for the names for the year and its divisions in the two languages have been formed quite independently of each other. Nevertheless the Italians appear to have already in their pre-Hellenic period advanced, if not to the arrangement of a fixed calendar, at any rate to the institution of two larger units of time. The simplifying of the reckoning according to lunar months by the of the decimal system, which was usual among the Romans, and the designation of a term of ten months as a "ring" (annus) or complete year, bear in them all the traces of a high antiquity. Later, but likewise at a period very early and apparently previous to the operation of Greek influences, the duodecimal system (as we have already stated) was developed in Italy, and, as it derived its very origin from the observation of the fact that the solar period was equal to twelve lunar periods, it was certainly applied primarily to the reckoning of time. This view accords with the fact that the individual names of the months (which can only have originated after the month was viewed as part of a solar year), particularly those of March and of May, were similar among the different branches of the Italian stock, while there was no similarity between the Italian names and the Greek. It is not improbable therefore that the problem of laying down a practical endar which should correspond at once to the moon and the sun—a problem comparable in some sense to the quadrature of the circle, and the solution of which was only recog-