< previous page page_231 next page >

Page 231
This scheme would produce an estimate of the solar year, which is about one day too long. This error could be corrected by omitting an intercalation every 22 or 24 years.
I do not suggest that the Romans followed exactly such a plan. It is known that they followed the general scheme of lunar years with alternating intercalations of a 27-day month and a truncated Februarius. And it is known that the calendar contained some error, since the office of the pontifex was deemed necessary.
9.1.2.4
The Intercalation Scheme of Macrobius
In this context, I will now examine Macrobius's statements regarding intercalation. Macrobius, of course, states that these rules apply to the calendar of Numa. According to Macrobius, intercalation was originally performed by rule, such that 22 or 23 days were intercalated in every second year, 90 days were intercalated at the end of eight years (two quadrennia), but only 66 days intercalated in the next two quadrennia. Macrobius does not state how the days were arranged, but you may surmise from the numbers that the patterns for the two quadrennia were {22, 23, 22, 23} and {22, 22, 22, 0}. This scheme would produce the following arrangement:
Year Number
Days in Year
Days Added
Total
1
355
0
0
2
377
22
22
3
355
0
22
4
378
23
45
5
355
0
45
6
377
22
67
7
355
0
67
8
378
23
90
9
355
0
0
10
377
22
22
11
355
0
22
12
377
22
44
13
355
0
44
14
377
22
66
15
355
0
66
16
355
0
66

 
< previous page page_231 next page >