5 For who can concede to them that the year has three hundred and sixty-five days only, in order that there may be twelve months of thirty days each, after the type of the twelve Æons, when the type is in fact altogether out of harmony [with the antitype]? For, in the one case, each of the Æons is a thirtieth part of the entire Pleroma, while in the other they declare that a month is the twelfth part of a year. If, indeed, the year were divided into thirty parts, and the month into twelve, then a fitting type might be regarded as having been found for their fictitious system.
But, on the contrary, as the case really stands, their Pleroma is divided into thirty parts, and a portion of it into twelve; while again the whole year is divided into twelve parts, and a certain portion of it into thirty. The Saviour therefore acted unwisely in constituting the month a type of the entire Pleroma, but the year a type only of that Duodecad which exists in the Pleroma; for it was more fitting to divide the year into thirty parts, even as the whole Pleroma is divided, but the month into twelve, just as the Æons are in their Pleroma.
Moreover, they divide the entire Pleroma into three portions—namely, into an Ogdoad, a Decad, and a Duodecad. But our year is divided into four parts, — namely, spring, summer, autumn, and winter. And again, not even do the months, which they maintain to be a type of the Triacontad, consist precisely of thirty days, but some have more and some less, inasmuch as five days remain to them as an overplus. The day, too, does not always consist precisely of twelve hours, but rises from nine to fifteen, and then falls again from fifteen to nine.
It cannot therefore be held that months of thirty days each were so formed for the sake of [typifying] the Æons; for, in that case, they would have consisted precisely of thirty days: nor, again, the days of these months, that by means of twelve hours they might symbolize the twelve Æons; for, in that case, they would always have consisted precisely of twelve hours.
Source: Adversus haereses (New Advent)