On which Gellius remarks:
"From these words it may be observed that the arrangement of (birth) days was such, that to any person born after sunset, and before midnight, the day from which that night had proceeded should be the birthday; but to any person born during the last six hours of the night, the day which should succeed that night must be the birthday."
This explanation might seem almost purposely written in reply to some such difficulty as occurred to Professor de Morgan (antè, p. 250.), when he remarks that, if birthday were to be confined to daylight, "a child not born by daylight would have no birthday at all!" But since it was notorious amongst the Romans that the civil day began at midnight, such a quæri solitum as this could never have been mooted, if the birthday observance had not been known and acknowledged to have a different commencement. In continuation of the same subject, Gellius proceeds to quote another passage from Varro, which I shall also repeat, not only as furnishing still farther proof that the Romans did not regard the night as forming any part of the birthday, but also as affording an opportunity of recording an opinion as to the interpretation of Varro's words, which, in this passage, do not appear to have ever been properly understood.
After stating that many persons in Umbria reckon from noon to noon as one and the same day, Varro remarks:
"Quod quidem nimis absurdum est; nam qui calendarum hora sexta natus est apud Umbros, dies ejus natalis videri debebit et calendarum dimidiatus, et qui est post calendas dies ante horam ejusdem diei sextam."
Now why should beginning one's birthday at noon appear so absurd to Varro? Simply because the hours of the night were not then supposed to be included in the birthday at all, and therefore Varro could not realize the idea of a birthday continued through the night.
He says that, according to the Umbrian reckoning, a person born on any day after the point of noon, would have only half a birthday on that day; and for the other half, he would have to take the forenoon of the following day. Varro had no notion of joining the afternoon of one day to the forenoon of another, because he looked upon the unbroken presence of the sun as the very essence of a natal day.
Nothing can be plainer than that this was the true nature of the absurdity alluded to; but it would not suit the prejudices of the commentators, because it would compel them to admit that sexta hora must have been in the afternoon, in opposition to their favourite dogma that it was always in the forenoon.
For if Varro had intended to represent sexta hora in the forenoon, he would have said that the other half-day must be taken from the afternoon of the pridie, instead of saying, as he does say, that it must be taken from the forenoon of the postridie of the Calends.
Consequently, Varro means by "qui Calendarum hora sexta natus est," a person born in the sixth hour of the day of the Calends; the sixth hour being that which immediately succeeded noon—the media hora of Ovid. But what Varro more immediately means by it is, not any particular point of time, but generally any time after noon on the day of the Calends.