"...as time zero today is still regarded the time of the birth of Christ, but this - according to the church - happened on Christmas Day and certainly not on our modern New Year's Day. In the 16th century in the Scandinavian countries they still regarded Christmas Day as the first day of the new year, and the same was valid in many places in Germany, while in Western Europe they regarded the birth of Christ as a less epoch-making happening than the moment of his conception on Lady Day. 

In the Frankish kingdom of Charlemagne and in the mediæval England they therefore counted the new year from the 25th of March, but in the Byzantine empire, which was playing a much greater role politically and culturally in the history of mediæval Europe than we nowadays currently usually think of, the new year started on the 1st of September, and this calendar was followed by the Roman popes for a fairly long time. 

In different places in Europe the 24th of September as well as the moving Easter Sunday were also at certain times regarded as the day a new year started. All this confusion, causing hopeless tasks for the historians trying to date events in the Dark Ages, was ultimately based on the antipathy of Old Christian Church towards the heathen revelries which in antiquity came to pass in conjunction with the Roman new year at the 1st of January.

The memory of this was never completely eradicated, however, and during the Renaissance when the arrangments of antiquity were esteemed again, even by the Church, the 1st of January was officially revivified as the first day of the new year.

Definitely established was this in the year 1588, when the pope Gregorius XIII sanctioned the calendar which after him is called the Gregorian and which now is the foundation of the almanac in practically all civilized countries.

That it is the Roman calendar of antiquity which ultimately is the foundation of our modern way to count years is most clearly seen in the curious fact that the short February has been retained though it would have been possible to let the whole of the Gregorian year have months with durations alternating between 31 and 30 days.

Fortunately Gregory XIII did not sanction the Roman way of naming dates, in which case his calendar hardly had been longlived. The Roman way of naming dates appears, to the posterity, as impractical as their numerals and in fact they systematically used in their calendar the same method of counting backwards as, to a lesser degree, was used in their system of numerals.

The Romans defined three fixed points in each month, which they called Kalendae, Nonae and Idus, and of these Kalendae always was on the first day, while Nonae and Idus not were placed at the same points in every month.

Modern date

March, May, July, October

January, August, December

April, June, September, November

February

1

Kalendis

Kalendis

Kalendis

Kalendis

2

VI. ante Nonas

IV. ante Nonas

IV. ante Nonas

IV ante Nonas

3

V. ante Nonas

III. ante Nonas

III. ante Nonas

III. ante Nonas

4

IV. ante Nonas

Pridie Nonas

Pridie Nonas

Pridie Nonas

5

III. ante Nonas

Nonis

Nonis

Nonis

6

Pridie Nonas

VIII. ante Idus

VIII. ante Idus

VIII. ante Idus

7

Nonis

VII. ante Idus

VII. ante Idus

VII. ante Idus

8

VIII. ante Idus

VI. ante Idus

VI. ante Idus

VI. ante Idus

9

VII. ante Idus

V. ante Idus

V. ante Idus

V. ante Idus

10

VI. ante Idus

IV. ante Idus

IV. ante Idus

IV. ante Idus

11

V. ante Idus

III. ante Idus

III. ante Idus

III. ante Idus

12

IV. ante Idus

Pridie Idus

Pridie Idus

Pridie Idus

13

III. ante Idus

Idibus

Idibus

Idibus

14

Pridie Idus

XIX. ante Kalendas

XVIII. ante Kalendas

XVII. ante Kalendas

15

Idibus

XVIII. ante Kalendas

XVII. ante Kalendas

XVI. ante Kalendas

16

XVII. ante Kalendas

XVII. ante Kalendas

XVI. ante Kalendas

XV. ante Kalendas

17

XVI. ante Kalendas

XVI. ante Kalendas

XV. ante Kalendas

XIV. ante Kalendas

18

XV. ante Kalendas

XV. ante Kalendas

XIV. ante Kalendas

XIII. ante Kalendas

19

XIV. ante Kalendas

XIV. ante Kalendas

XIII. ante Kalendas

XII. ante Kalendas

20

XIII. ante Kalendas

XIII. ante Kalendas

XII. ante Kalendas

XI. ante Kalendas

21

XII. ante Kalendas

XII. ante Kalendas

XI. ante Kalendas

IX. ante Kalendas

22

XI. ante Kalendas

XI. ante Kalendas

IX. ante Kalendas

VIII. ante Kalendas

23

IX. ante Kalendas

IX. ante Kalendas

VIII. ante Kalendas

VII. ante Kalendas

24

VIII. ante Kalendas

VIII. ante Kalendas

VII. ante Kalendas

VI. ante Kalendas

25

VII. ante Kalendas

VII. ante Kalendas

VI. ante Kalendas

V. ante Kalendas

26

VI. ante Kalendas

VI. ante Kalendas

V. ante Kalendas

IV. ante Kalendas

27

V. ante Kalendas

V. ante Kalendas

IV. ante Kalendas

III. ante Kalendas

28

IV. ante Kalendas

IV. ante Kalendas

III. ante Kalendas

II. ante Kalendas

29

III. ante Kalendas

III. ante Kalendas

II. ante Kalendas

Pridie Kalendas

30

II. ante Kalendas

II. ante Kalendas

Pridie Kalendas

 

31

Pridie Kalendas

Pridie Kalendas

 

 

The three words in second casus are respective Kalendis and Kalendas, Idibus, Nonis and Nonas. Pridie means the day before.

It all is quite strange. The days of the month were numbered in relation to a coming date, and more than half the month was counted backwards from the first in the coming month. 

Linguistically it was further complicated, from a modern view, by the fact that the names of the months in Latin not are nouns, but adjectives and that Kalendae, Nonae and Idus are plural forms. Idibus Martiis, which undoubtedly is the most known of all Latin dates, consequently means something like 'by the Marchian Ides'. The last mentioned word, Idibus or in nominative Idus, is a kind of proper name and as such of course not possible to translate; it had initially something to do with the full moon."

(Alf Henrikson "Antikens historier")