XVII. Be not ignorant of this, however, that those four determining periods, 1192 which we have mentioned, although they are approximated to the Kalends of the following months, yet hold each the middle of a season, viz., of spring and summer, and autumn and winter. And the beginnings of the seasons are not to be fixed at that point at which the Kalends of the month begin. But each season is to be begun in such way that the equinox divides the season of spring from its first day; and the season of summer is divided by the eighth day before the Kalends of July, and that of autumn by the eighth before the Kalends of October, and that of winter by the eighth before the Kalends of January in like manner. 1193 1192     Temporum confinia. 1193     [Compare what is said of Hippolytus, vol. v. p. 3, this series. See the valuable work of Professor Seabury on the Calendar, ed. 1872.]