These two dials were fixed on pillars behind the Rostra in the Forum, the most convenient place for regulating public business, and there they remained even in the time of Cicero[409].

But in the censorship next following that of Philippus the first water-clock was introduced; this indicated the hours both of day and night, and enabled every one to mark the exact time even on cloudy days[410].
Thus from the time of the Punic wars the city population reckoned time by hours, i.e.twelve divisions of the day; but as they continued to reckon the day from sunrise to sunset on the principle of the old agricultural practice, these twelve hours varied in length at different times of the year.

In mid-winter the hours were only about forty-four minutes in length, while at mid-summer they were about seventy-five, and they corresponded with ours only at the two equinoxes.[411] This, of course, made the construction of accurate dials and water-clocks a matter of considerable difficulty.

It is not necessary here to explain how the difficulties were overcome; the reader may be referred to the article "Horologium" in the _Dictionary of Antiquities_, and especially to the cuts there given of the dial found at Tusculum in 1761.[412] Sun-dials, once introduced with the proper reckoning for latitude, soon came into general use, and a considerable number still survive which have been found in Rome.

In a fragment of a comedy by an unknown author, ascribed to the last century B.C., Rome is described as "full of sun-dials,"[413] and many have been discovered in other Roman towns, including several at Pompeii.