YALE UNIVERSITY
LIBRARY

THE LIBRARY OF THE
DIVINITY SCHOOL

MEDIEVAL RECKONINGS OF TIME

HELPS FOR STUDENTS OF HISTORY. No. 3.
Edited by C. Johnson, M.A., F.S.A., H. W. V. Temperley, M.A.,
and J. P. Whitney, D.D., D.C.L.

MEDIEVAL RECKONINGS
OF TIME

BY
REGINALD L. POOLE
M.A., LL.D., LiTT.fr v
KEEPER OF THE ARCHIVES OF THE UNIVERSITY OF OXFORD
AND FELLOW OF MAGDALEN COLLEGE AND OF THE BRITISH ACADEMY

LONDON
SOCIETY FOR PROMOTING
CHRISTIAN KNO W L&JIG^F
NEW YORK: THE MACMILL^N COMPAQ
1921

First published . 1918
Reprinted . . . 1921

CONTENTS

n.
MONTHS -

PAGE

ELEMENTS  . 7
I.
DAYS  u
Holy Days - - . . . - 12
The Days of the Week - - . - 17
Rule to find the day of the week - - 19

21

The Days of the Month - - - - 23
Seasons - - - - - - 25
III.
YEARS ..... .27
Cycles of Years - - - - - 29
Easter Tables - - - . - 32
Eras - - - - - . .36
The Christian Year - - - 39
The Beginning of the Year - - - 41

MEDIEVAL RECKONINGS
OF TIME
ELEMENTS
In the following sketch an attempt is made to
describe in the simplest terms the way in which
time was reckoned in the Middle Ages. Ac
cordingly we have only to do with the Julian
Calendar, or Old Style, and are not concerned with
the reformed Calendar, or New Style, which was
introduced by Pope Gregory XIII. in 1582. We
are also under the domination of the Ptolemaic
system, which supposes that the Sun and the planets
revolve round the Earth ; and this belief is so deeply
embedded in language that it is natural to use forms
of statement which assume it. Again, when we
speak of the Day and the Year, we mean the
periods known by common observation, and not
other periods arrived at by the calculation of
astronomers. Nor do we aim at precision in our
7

8 MEDIEVAL RECKONINGS OF TIME
figures: for our purpose the Solar Year consists
of 365 \ days, not of 365 days 5 hours 48 minutes
46 seconds.
In an outline like the present it has been judged
most practical to exclude the discussion of Oriental
or Byzantine chronology; still less do we propose
to say anything about the system, based upon the
Lunar Year, which prevailed and still prevails
among Saracens.
Fuller information will be found in :
Ducange, "Glossarium mediae et inflmse Latinitatis," ed.
G. A. L. Henschel, 7 volumes, 1840-1850 (under Annus,
Men sis, etc.).
" Art de Verifier Ies Dates," ed. Saint-AUais, Part II. , Vols.
I., IL, 1818.
H. Grotefend, " Zeitrechnung des Deutsehen MIttelalters und
der Neuzeit," Vol. I., 1891; and Vol. IL, Part IL, 1898.
A. Giry, " Manuel de Diplomatique," Book IL, 1894,.
F. Ruhl, " Chronologie des Mittelalters und der Neuzeit," 1897.
F. K. Ginzel, " Handbuch der mathematischen und tech-
nischen Chronologie," Vol. III., 1914.
For common use the following handbooks may
be mentioned:
Sir Harris Nicolas, " The Chronology of History," 1833; new
edition, 1840.
Augustus De Morgan, " The Book of Almanacs," 1851 ; 8rd
edition, 1907.
J. J. Bond, " Handy-Book of Rules and Tables for verifying
Dates with the Christian Era," 1875; 4th edition, 1889.
H. Grotefend, " Taschenbuch der Zeitrechnung," 1898.
E. A. Fry, " Almanacks for Students of English History,"
1915.

ELEMENTS 9
Time has always been reckoned by the move
ments of the Earth, the Sun, and the Moon. The
Earth rotates upon its axis in a Day, which we
divide into 24 hours. The Sun was deemed to
revolve round the Earth in 365 \ days : this we call
a Solar Year of 365 days, and provide for the odd
quarter by intercalating a day in the February of
every fourth year (Leap Year). The year is divided
for convenience into twelve Months.
Another fact which has always been brought
prominently before men's minds, especially in
times when husbandry was their principal occupa
tion, is that the Moon revolves round the Earth
in about 29| days. It was obviously impossible
to combine this Lunar Month even approximately
with the number of days in the Solar Year, and
infinite ingenuity was exercised in order to har
monize the discordant elements. The four quarters
or Phases of the Moon were very roughly repre
sented by the Week of seven days. This was
derived in ancient times from the East, and was
firmly established in the Calendar through the
observances of the Christian Church. It is an
independent form of reckoning which stands in
no exact relation to the Month or Year, whether
Solar or Lunar. The days of the Week go on
continuously without regard to the larger units.

10 MEDIEVAL RECKONINGS OF TIME
Hence, if a complete chronological scheme was to
be constructed, there was a third element to be taken
into account; and to bring the movements of the
Sun and the Moon into agreement with the days of
the Week introduced a further problem, which
was of paramount importance in the Middle Ages,
because on its solution depended the date of
Easter. To this we shall return at a later stage.

DAYS
The Day was divided into two periods, each of
twelve hours, beginning respectively at sunset
and sunrise. Hence in the different seasons the
hours varied greatly in length. At midwinter an
hour in the daytime was reduced to 40 minutes,
and an hour at night extended to 80 minutes.
But the division into 24 hours of equal length was
not unknown; we may find it noted in calendars,
but it does not seem to have been used in practice.
Down to the fourteenth century the length of the
hours changed with the seasons, and these unequal
hours were numbered from sunset and sunrise.
It was the placing of clocks with bells in churches
and other public buildings that gradually estab
lished a system of uniform hours. We still mark
its origin when we speak of the hour of the clock,
" o'clock." These hours were generally numbered
from 1 to 12, in two series, beginning at midnight
and noon respectively. But in Italy and Germany
the numbering from 1 to 24 was long retained. In
n

12 MEDffiVAI RECKONINGS OF TIME
some places, particularly in Italy, though the hours
were made equal, the clocks were so far adapted
to the local custom that the beginning of the day
(24 o'clock) was altered from season to season, and
was fixed at or soon after sunset. In Italy this
practice was not given up until the opening of the
nineteenth century.
The hours of Church service followed the system
of variable hours: Mattins before dawn, Prime at
sunrise, Tierce at 3, Sext at 6, Nones at 9, Vespers
or Evensong at 11, reckoned from sunrise, and
Compline after sunset. But in course of time
Nones were set back to midday, whence we call
it "noon." It should be noted that in ecclesi
astical use the day begins with Evensong of the
preceding day.

HOLY DAYS
Holy Days occur either on fixed days of the
month or on days determined by the varying date
of Easter. We speak, therefore, of Fixed and
Moveable Feasts, and use the term Feast (as it is
used in the Prayer-Book) to denote any day of
special religious observance, whether it is a Feast
or Fast.
Fixed Feasts are those commonly known as

HOLY DAYS 13
Saints' Days. They are the commemorations found
in the Roman calendar, increased as time went
on by feasts observed either generally or in particu
lar countries or in particular localities. The day,
unless specially distinguished, as the Nativity of
St. John Baptist or the Conversion of St. Paul, is
understood to be that of the Saint's burial (depo-
sitio). But it sometimes happened that the Saint's
body was removed to a different place, and this
Translation commonly took place at a time of year
suitable for a great popular commemoration.
For example, the Feast of St. Martin was Novem
ber 11 (S. Martin in hieme), but his Translation on
July 4. The Martyrdom of St. Thomas was on
December 29, but his Translation on July 7. It
should be noticed that all Feasts were not every
where kept on the same day, and in cases of doubt
it is necessary to consult a book of reference.
Sometimes also there is a risk of confusion through
the Saint's title being omitted: thus St. Stephen
the Martyr is commemorated on December 26,
but St. Stephen the Pope on August 2. In Eng
land the neighbouring Feasts of St. Edmund the
Archbishop on November 16 and of St. Edmund
King and Martyr on November 20 are not always
properly distinguished.
Moveable Feasts may vary, according to the

14 MEDIEVAL RECKONINGS OF TIME
date of Easter, as much as five weeks; for Easter
may fall on any Sunday between March 22 and
April 25 inclusive. The earliest of these Feasts is
Septuagesima Sunday, which may occur as early as
January 18, though it has so occurred less than
once in a century. The period during which this dis
turbance lasts continues strictly until the Sunday
next but one before Advent — that is, until late in
November. But in practice it formerly ended at
Whitsunday, for the usage of numbering the
Sundays between that Feast and Advent was only
introduced by degrees. The Sunday after Whit
sunday was kept, perhaps first in England, as the
Feast of the Holy Trinity at least from the eleventh
century, and the Thursday following was estab
lished by Urban IV. in the thirteenth as the
Feast of Corpus Christi. When the Sundays came
to be numbered, they were counted, as they still
are in the Roman use, after Whitsunday; but in
England the practice of reckoning them from Trinity
prevailed. In other parts of the year, too, it was
common to speak of Sundays not by their place
in the Church year, but with the opening word of
the Introit in the Mass for the day. Some of these
designations are particularly often used, and indeed
hold their ground in countries of the Lutheran
confession down to the present day :

HOLY DAYS 15
Septuagesima  Circumdederunt
Sexagesima . . . . . . . . Exsurge, Domine
Quinquagesima . . . . . . Esto mihi
First Sunday in Leut . . . . Invocavit
Second Sunday in Lent . . Reminiscere
Third Sunday in Lent . . . . Oculi
Fourth Sunday in Lent . . . . Lcetare, Jerusalem
First Sunday after Easter . . . . Quasi modo geniti
Second Sunday after Easter . . Misericordia Domini
Third Sunday after Easter . . . . Jubilate
Fourth Sunday after Easter . . Cantate
Fifth Sunday after Easter . . . . Vocem jocundifatis
The chief Feasts had a subsidiary observance
a week later : this was called the Octave, in English
the Utas. It was very commonly used in sum
monses to councils, parliaments, and the like, as
a convenient date for attendance when the great
festival was over. For this purpose also the
Quindene or Quinzaine, a fortnight after the Feast,
was sometimes appointed.
It was not usual until about the thirteenth
century to date letters or other documents by the
Holy Days, but chroniclers constantly recorded
events as occurring on particular festivals. They
were also found convenient for specifying the dates
on which rents should be paid; the festival served
as the most practical reminder of the obligation
at a time when almanacks were not in popular use.
Hence we constantly read in charters of the grant
of land or of a house subject to the payment of a

16 MEDIEVAL RECKONINGS OF TIME
definite rent, not merely at the four customary
quarter-days (all of which were feasts), but also,
for example, twice a year on the feasts of the Con
version of St. Paul and of St. James, January 25
and July 25. This arrangement was the easier
to make, since in six months of the year a great
feast fell on the 8th of the Kalends, and in two
others on the 7th or 9th. One date which was
very often used in England as a term for paying
rent or collecting dues was, unlike those which have
been mentioned, connected with a moveable feast.
This was Hock Day or Hoke Day, on the Tuesday
after the Octave of Easter. Hocktide included
that day and the Monday preceding it. It was
taken as a convenient time for spring payments,
when the autumn payments were at Michaelmas;
just as the Exchequer opened its two annual
sessions on Hock Monday and on the morrow of
Michaelmas.

THE DAYS OF THE WEEK 17

THE DAYS OF THE WEEK
The days of the week were named from the seven
heavenly bodies known to the ancients: the Sun,
Moon, Mars, Mercury, Jupiter, Venus, and Saturn.
English and Dutch are almost the only languages
in which all the seven are still used, the classical
names of deities being rendered by their presumed
equivalents in Teutonic mythology. Everywhere
else one or more of them has been superseded.
The Sabbath came in first (this is the only
change found regularly in German); then the
Lord's Day took the place of the dies solis. Thus
we get the series of names as they stand in Latin,
French, Italian, and Spanish. The first day of
the week was dominica or dies dominicus ; the last,
sabbatum or dies sabbati. The Sabbath in the
Middle Ages was never any day but Saturday; its
occasional transference to Sunday rested on a
theory not earlier than the Reformation, and to call
Saturday dies Satumi is a modern artificial revival-
A day of the week was known as a feria, and the
ferice were numbered from 1 to 7; but feria prima
for Sunday and feria septima for Saturday were
hardly ever used. Hence the days of the week might
be designated, as they still are in the Roman Church

18 MEDIEVAL RECKONINGS OF TIME
Calendar, dies dominions, feria secunda, feria tertia,
feria quarta, feria quinta, feria sexta, sabbatum.
These names are preserved in a translated form in
Portuguese down to the present time.
The numbers of the days in each week were
marked in calendars throughout the year by the
letters of the alphabet from A to G. These are
the Ferial or Day Letters, the letter of January 1
being A. Every day in the year has its letter, and
when we know the day of the week of January 1,
we can reckon from it to any other day. For
instance, if A, January 1, is Sunday, then February 1,
D, is Wednesday. This method, we shall see,
was used in order to find the week day of the 1st
of every month. But the Day Letters are chiefly em
ployed for a different purpose — namely, to note
the place occupied by Sunday in the first week of
the year. If A, the first day, is Friday, then the
Dominical or Sunday Letter, as it is called, is C;
if the first day is Saturday, the Sunday Letter is
B. Consequently, while the Day Letters run
forward from A to G, the Sunday Letters go back
wards in successive years in the order A, G, F, E,
D, C, B. A Leap Year has two Sunday Letters,
one {e.g., F) down to February 24, the second {e.g.,
E) for the rest of the year.
A peculiarly complicated method of indicating

THE DAYS OF THE WEEK 19
each day of the year is that known as the Cisio-
janus. This is a series of twenty -four hexameters
of execrable Latinity, made up chiefly from frag
ments of the names for the Holy Days. There are
various forms of it, one of which begins, for the
month of January :
Cisio Janus Epi sibi vendicat Oc Feli Mar An,
Prisca Fab Ag Vincen Ti Pau Po nobile lumen.
Each syllable denotes a day, and, as the same
syllable occurs in more than one month, it is
necessary to quote several syllables or else to
mention the month as well as the syllable.
Happily this barbarous invention, though common
in Germany, is hardly ever found in England
except as an adjunct to calendars.
Rule to Find the Day of the Week
The day of the week on which any historical
event occurred can easily be ascertained from the
tables given in the books named on page 8. But
it is a good thing to have a simple formula which
will bring out the required result during the period
of the Julian Calendar without the need of con
sulting books. For this purpose a most practical
formula was devised by Father Chambeau, S.J., in
1892.* We add together five numbers and divide
* See " Bibliotheque de l'Ecole des Chartes," liv. 595.

20 MEDIEVAL RECKONINGS OF TIME
the sum by 7: the remainder gives the day of the
week, 1 being Sunday, 2 Monday, and so on. The
numbers are as follows:
1. The Day of the month.
2. The Month number — i.e., the ferial letter
giving the weekday of the 1st of the month:
1, 4, 4, 0, 2, 5, 0, 3, 6, 1, 4, 6, as printed below,
p. 22, but with 0 in place of 7.
3. The Year in the Century.
4. A quarter of this, omitting fractions.
5. The Hundred number, which is 18 minus the
number of the hundreds (not centuries,
because we begin with 0, not with 1).
The formula then is: The day of the week is the
remainder of d+m + y+% + {18-h)
7
The operation may be simplified by casting out
sevens at every stage. Thus, Thomas Becket was
consecrated on June 3, 1162.
3 or 3
5 5
62 6
15 1
J_ _0
7) 92 7) 15
13 • 1 = Sunday. 2 • 1 =Sunday .

II
MONTHS
The Solar Year is arbitrarily divided into twelve
Months. Our system of months is derived from a
year which began on the 1st of March, as is shown
by their names from September to December.
In the calendar as revised by Julius Caesar
the months were given alternately 31 and 30 days
(except that in two cases two months of 31 days
succeeded one another), and the last month,
February, was left defective by two days. The
beginning of the Year was fixed in January, but
the names of the months indicating a different
commencement were not altered, except that
Quintilis was shortly afterwards changed to Julius
and Sextilis to Augustus.
As every month, excepting February, was two
or three days in excess of four weeks, a series of
twelve numbers was arranged to show the weekday
of each; 1 being Sunday, 2 Monday, and so on.
These numbers were: 21

22 MEDIEVAL RECKONINGS OF TIME
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
255136147257
These were employed to assist the computation
of Easter, and were called the Solar Regulars.
But in practice it is convenient to reduce these
figures by 1, so as to make the numbers agree
with the Ferial Letters, and 1=A the first day of
the year; 7 and 0 are of course for this purpose
exchangeable. We take, therefore, the following
series: Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
144725736146
These were remembered by means of verses. Thus
in the thirteenth century,
Altitonans Domiiuis Divina Gerens Bonus Extat:
Gratuito Celi Fert Aurea Dona Fideli.
or, AltaDomat Dominus; Gratis Beat Equa Gerentes:
Contemnit Fictos, Augebit Dona Fideli.
In English we have an old piece of doggerel :
At Dover Dwell George Brown Esquire,
Good Christopher Finch, And David Friar.
In Leap Year the numbers for the time from
January 1 to February 24 have to be diminished by
one.

THE DAYS OF THE MONTH 23

THE DAYS OF THE MONTH
The old Roman way of reckoning the days from
, three fixed points in the month never fell altogether
out of use. This system counted backwards from
the Kalends or 1st of the following month, the
Ides on the 13th of the month, and the Nones on
the 5th, including both the extreme days in the
calculation. In March, May, July, and October
the Nones and Ides were placed two days later.
In Leap Year the leap occurred after the 24th of
February: the intercalated day, bissextum or dies
bissextus, was inserted before the 6th of the Kalends
of March, which in common years was 24th of
February. Hence in Leap Year any observance
belonging to the 24th of February was kept on the
25th. This happened in regard to the Feast of
St. Matthias. Mistakes sometimes arose through
the scribe, from familiarity with the modern way
of counting the days of the month, naming the
Kalends of the current month instead of those of
the following month. A date before the Kalends
of January might even be understood as falling
in a wrong year. For instance, the death of
William Longsword, Duke of Normandy, was
recorded as taking place anno ab incarnationc

24 MEDIEVAL RECKONINGS OF TIME
Domini DCCCCXLIII, XVI Kalendas Januarii.
This statement was supposed by the annalist
Flodoard of Rheims to refer to the year 943, but
in fact it means the 16th day before January 1, 943.
The modern way of counting the days straight
on from the 1st to the last day of the month came
to us from Syria and Egypt; but in the West it is
hardly found before the second half of the sixth
century. The practical instinct of Pope Gregory
the Great led him to introduce it into his chancery,
but his successors went back to the older style.
The modern plan, however, won more and more
acceptance among the Lombards and Franks,
and when people wrote in the vernacular it was
regularly employed. It is worth noting, on the
other hand, that the revival of learning under
Charles the Great led to an official reversion to the
classical system, which long held its ground in the
Imperial chancery.
A peculiar method of reckoning the days of
the month is known as the Custom of Bologna
(consuetudo Bononiensis). This counted the days
forwards from 1 to 16 (or 15 in months of thirty
days), and backwards from the last day in the
second half of the month. The day in the one case
was noted as intrante mense, in the other as exeunte
or stante mense. The origin of this system is not

THE DAYS OF THE MONTH 25
clear, but it is unconnected with the old Roman
method, because that reckoned back, not from the
last day of the month, but from the 1st of the fol
lowing month. The consuetudo Bononiensis was
not often used outside Italy, except when Italian
notaries were employed in the Imperial chancery;
but from the twelfth century it is occasionally
found in France, probably as a consequence of the
ascendancy of the law school of Bologna, which
was frequented by men who intended to enter the
profession of notaries. That it did not pass into
England so early is explained by the fact that it
was not until a later time that notaries were
employed here; but from the fourteenth century
we may find it in English documents drawn up by
Papal notaries. The Bologna reckoning survived
on the Continent as late as the fifteenth century,
but before that date it was generally passing out of
fashion. Seasons
The year was popularly divided into four Seasons,
the middle of each of which was fixed at the time
of the Equinoxes and Solstices. The dates at
which they began are variously stated. Strictly,
of course, they should be placed as nearly as
possible halfway between the Equinox and the

26 MEDIEVAL RECKONINGS OF TIME
Solstice; and so Bede made Spring begin on Feb
ruary 7, Summer on May 9, Autumn on August 7,
and Winter on November 7 — that is, in every case
on the 7th of the Ides of the month. But he
noticed also that Isidore of Seville put the dates
more than a fortnight later: on February 22,
May 24, August 23, and November 23.* In medieval
calendars both dates are sometimes entered; but in
time the Isidorian system prevailed, though two
of the dates were shifted one day. Hence, at
least in the later Middle Ages, the seasons were
taken as beginning on the Feasts of St. Peter in
Cathedra (February 22), St. Urban (May 25), St.
Bartholomew (August 24), and St. Clement (Novem
ber 23).
* " De Temporum Ratione," xxxiii.

Ill
YEARS
A year may be indicated by its place in the current
reign, or in a Cycle or repeating series of years,
or in a series counted from a definite date or Era.
It should be premised that the ancient eras, such
as the Year of the World (which was diversely
computed), the Olympiad, the Year of Rome, and
the Seleucid Era, may be here left out of account,
though they are occasionally recorded in chrono
logical tables. For practical purposes we may
limit ourselves to the three modes of dating men
tioned above.
Under the Roman Empire the year was marked
by the names of the Consuls, and as their office was
annual their names supplied a precise date. But
no Consuls were appointed after a.d. 534 in the
West and 541 in the East, and the years following
were styled post consulatum Paulini in the West
and Basilii in the East. Then in 566 the Emperor
Justin II. assumed the Consulship himself, and from
that time the Imperial years were denoted in such
27

28 MEDIEVAL RECKONINGS OF TIME
a form as anno II. post consulatum Mauricii, the
Consulate being usually deemed to begin on Jan
uary 1 following the Emperor's accession. Thus
the post-consulate reckoning acquired the same
meaning as that by the years of the Sovereign,
which had been used in very ancient times and
which is continued in monarchical countries to
the present day. But the date from which these
Regnal Years were counted was not everywhere the
same. Among the Franks, for example, they began
with the King's accession; but when Charles the
Great became Emperor in 800, he dated from his
coronation, and this practice was usually main
tained by the French Kings. In Germany the
Kings as a rule dated their years from their acces
sion, but if they became Emperors from their
coronation, or they might give both dates; the
rule, however, is not free from exceptions. In
England the year was always reckoned from the
King's coronation until the death of Henry III.,
when, as his son was absent on crusade, a change
had to be devised; and from that time the Kings'
years have been counted from their accession.
The Popes have almost always reckoned their
years, Pontifical Years, not from their election, but
from their ordination or consecration.

CYCLES OF YEARS 29

CYCLES OF YEARS
The first Cycle which was used for historical
purposes is that of the Indiction, which consisted
of fifteen years. This Cycle is computed from
September 1, a.d. 312, though it is probable that
it was, in fact, instituted for fiscal purposes in 297.
But the date from which the first Indiction was
reckoned is immaterial, for the number of the
Cycle is hardly ever spoken of. All that is men
tioned is the place which a particular year holds
in an unspecified Cycle of fifteen years. We can
only tell which that Cycle is by comparing other
dates named in connection with that of the Indic
tion — for instance, the post-consular or regnal
year, or later on the year of the Incarnation. The
Indiction, therefore, is mainly useful to us now as
a means of checking and correcting other chrono
logical elements which accompany it. The date was
always expressed in such a form as prima Indictione,
not in the first year of the Indiction. When we wish
to translate it into terms of years of the Christian
Era, it is the rule to denote it by the year in which
the greater part of it falls, so that the first Indiction
is called the Indiction of 313.
This is worth remark because the fact has not

30 MEDIEVAL RECKONINGS OF TIME
always been clearly stated; indeed, a famous
scholar, Kemble, treated at large upon the chrono
logy of Anglo-Saxon charters, in which it is a normal
feature, under the belief that the Indiction began
in the September of what we should call the
current year. Such a practice is unknown except
at Genoa.
The Indiction, however, did not always start
from the same day. Besides the original, Greek
or Constantinopolitan, Indiction of September 1,
two varying modes of reckoning were developed.
One of these, called the Cesarean or Imperial
Indiction, or the Indiction of Constantine, began
on September 24; this was invented probably in
order to bring it into accord with the autumnal
equinox. It won favour through the authority
of Bede, who adopted it; and it is conveniently
distinguished as the Bedan Indiction. The other
is known as the Roman or Pontifical Indiction;
this began on December 25, or perhaps sometimes
on January 1, and was no doubt devised in order
to make the Indiction correspond with the common
reckoning of the year. Of these three varieties
the Greek Indiction was regularly employed by
the Popes until 1087; after this the usage fluctu
ates until under Alexander III. (1159-1181) the
Bedan Indiction of September 24 became estab-

CYCLES OF YEARS 31
lished. From Italy the Indiction passed into the
Imperial chancery with the coronation of Charles
the Great in 800; first the Greek Indiction,
and then alternatively1 the other two modes of
reckoning. In England the Indiction of Bede was usual, but
the simpler computation from Christmas gradually
came to be employed. Outside the Papal and
Imperial chanceries the Indiction was mainly an
ecclesiastical system. It may be found in charters
for churches, and historians occasionally make
mention of it; but it passed more and more out of
general use. Only the notaries public who received
their authority from the Pope or Emperor, and who
delighted to amplify their certificates in the most
precise fashion, never omitted to insert the Indic
tion in their official Acts.
Other Cycles which were employed were invented
for the technical purpose of ascertaining the date
of Easter,, or more exactly of finding the fourteenth
day of the Moon of the Vernal Equinox. In what
relation Easter stood to that day had been a subject
of long dispute, but it was nearly settled by the
fifth or sixth century; and the definition which
became accepted was that Easter Day was the
Sunday next following the Full Moon on or next
after March 21.

32 MEDIEVAL RECKONINGS OF TIME

EASTER TABLES
In order to find the term from which Easter —
the governing date in the Christian Year — should
be computed, it was sought to reconcile the dis
cordant elements furnished by the Solar Year of
365^ days, the Lunar Year of 354 days, and the
unrelated march of the Days of the Week, by
constructing a Cycle, or series of years, at the end
of which the elements would start afresh as they
were at the beginning. To adjust the Week Days
to the Solar Year was an easy matter. If the Year
had consisted of just 365 days, the same Week Day
would be the first day of every seventh year; but
the insertion of a day in Leap Year makes it neces
sary to multiply the number by four, and thus we
get the Solar Cycle of 28 years.
To adjust this Cycle to the phenomena of the
Moon was much more complicated. At Rome it
was reckoned in the third century that a Cycle of
84 years — that is, of three Solar Cycles — would form
a period at the end of which the elements would
again be in harmony; but in fact, while it, of
course, brought the Year and the Day of the Week
into agreement, it involved an error of a day and
a quarter in the calculation of the Moon. It was

EASTER TABLES 33
such a Cycle that passed into use in the Celtic
Church and led to a long controversy with those
who adopted a more scientific system, which was
not appeased until some time after the Council
of Whitby in 664.
Evidently in order to attain anything like ac
curacy it was necessary also to take account of
the periods of revolution of the Moon. Now it
had been calculated in ancient Athens that 235
Lunar Months were almost exactly equal to 228
Solar Months, or 19 years, at the end of which the
two systems were again in agreement. This Lunar
Cycle was adopted in the East as one of the deter
mining data for the reckoning of Easter. The
Alexandrians constructed a Great Cycle, or Annus
Magnus of 532 years, the product of the 28 years
of the Solar multiplied by the 19 years of the Lunar
Cycle. This reconciled the three discordant ele
ments: when the Cycle ended, the same day of the
month, the same day of the week, and the same
phase of the moon occurred as at its beginning.
It should be observed that, when this system was
finally established, the Lunar Cycle was computed
from two different points of departure. That
which was known as the Lunar Cycle — the Greek
reckoning — started three years later than that
which became accepted in the W-esC^^The. latter

34 MEDIEVAL RECKONINGS OF TIME
is called the Cycle of 19 years, cir cuius decern-
novennalis, or the Golden Number, and the number
of any year in it is likewise spoken of as the Golden
Number. A table based on the period of 532 years was
constructed in 457 by Victorius of Aquitaine, and
was long used in Italy and longer still in Gaul. It
was superseded by a new Cycle of nineteen years
drawn up by Dionysius Exiguus in 525, but
perhaps not brought into currency until more than
a hundred years later. To this we shall return
when we come to consider the origin of the Year
of the Incarnation.
Although the Cycle of 84 years was superseded
by a more accurate system introduced from Alex
andria, one feature in it continued a permanent
element in the chronological system of the West.
This feature is the arrangement of the Lunar Years.
A Lunar Year, we have seen, consists of 354 days.
It was divided into twelve Lunar Months, or Luna
tions, arranged alternately as months of 30 and of
29 days (known respectively as Full and Hollow
Months, and named from the month in the
Solar Year in which they ended). Now, as the
Lunar Year is eleven days shorter than a Solar
Year, the Moon is eleven days older at the end of
the year than it was at the beginning. The age

EASTER TABLES 35
of the Moon on January 1 is called the Epacts
{adiectiones lunce). If the number is 1 in a given
year, it is 12 in the next and 23 in the third. The
year was reckoned for this purpose as beginning
on September 1 preceding the current year. When
the number reached 30, this number was deducted
from the Epacts and an additional or Embolismal
Month of 30 days was intercalated in the Year:
this month was inserted in seven years of the
Cycle of nineteen. In Leap Year also an inter
calated or Embolismal Day was added. But when
the Cycle was completed, the Lunar series was
found to be one day in advance of the Solar; so the
Moon was assumed to skip a day — this was called
the saltus Iwnhe — in other words, a month in the last
year of the Cycle was diminished by one day.
This skip was commonly made in November, but
sometimes, after the Alexandrian practice, in
July. It will be understood that the Moon computed
by these means is not the Moon as observed, but
what is called the Calendar Moon, which may or
may not agree with the true Moon. To take an
example within recent memory, in 1903 the Easter
tables, following the modern reformed Calendar,
placed the Full Moon of the Vernal Equinox
on April 11, and consequently Easter Day was

36 MEDIEVAL RECKONINGS OF TIME
kept on the 12th. But our almanacks tell us that
the Moon was not full until eighteen minutes after
midnight, therefore on April 12 ; so that, if we had
followed observation instead of computation, Easter
should have been kept on the 19th.
The Easter tables led primarily to the ascertain
ing of the Full Moon, the Paschal Term, as it was
called. To find the date of the next — that is,
Easter — Sunday, a series of numbers known as
Concurrents was arranged to indicate the Day
of the Week on March 24; these numbers, running
from 1 to 7 — 1 being Friday, 2 Saturday, etc.,
and a number being omitted each Leap Year —
were assigned in order to each year, and the year
for this purpose was reckoned from March 1. If
March 24, for instance, falls on a Sunday, the
Concurrent of that year is 1 and the Sunday
Letter F; March 1 is a Friday. This explains why
the numbering of the Concurrents starts from that
day of the week. ERAS
We have had to say something about the elements
of the Easter Cycle, because these data are not
infrequently noted in chronicles and in charters;
but they cannot be included among the common

ERAS 37
reckonings of time. Two of these Cycles, however,
acquired great importance because they developed
into Eras; that is to say, the first year in a table
intended for finding Easter came to be accepted
as an Era from which years might be counted
continuously for all time. For while the Cycles
were constructed for practical use, and therefore
started either from about the year when the com
piler wrote or else from the termination of the Cycle
then current, it was not unnatural that he should
make tables of a more ambitious kind extending
back to an earlier date, so that they might be
available for reference and for historical purposes.
If he did this, he would repeat the existing Cycle
for periods of the past and calculate them on the
same principle as that which served for the present
and future. To such amplified Cycles may be
traced in all probability the origin of the Spanish
Era and certainly that of the Year of the Incarna
tion. The Spanish Era makes its appearance among
the orthodox Christians in Spain soon after the
middle of the fifth century; from them it was
adopted by the Arian Visigoths. At that time the
old-established Easter table used at Rome was a
Cycle of 84 years (or three Solar Cycles) beginning
on January 1, 298. It was just about to be revised,

38 MEDIEVAL RECKONINGS OF TIME
but it is unlikely that the change became known
in Spain immediately. In that country, it seems,
a chronologer drew up a series of cycles to precede
that of 298 and to include the presumed date of
the Resurrection; and calculating thus backwards,
he made his first cycle begin on January 1, 38 B.C.
This theory presumes that he accidentally added
in 298, which was in fact the first year of the next
cycle. It is an interesting coincidence that if he
had performed the same process with the Oriental
cycle of Cyril of Alexandria, which was of 95
years (or five Lunar Cycles), and made the same
mistake, he would have arrived at precisely the
same starting-point. And the same result would
have been obtained had he worked back from the
earlier table of Theophilus of Alexandria, which
began in 380 : deducting 22 Lunar Cycles or
418 years from 380, he would have come to
38 B.C.
The Spanish Era is peculiar to the Peninsula
and to the south-west parts of Gaul (Aquitaine and
Languedoc), which were for a time subject to the
Visigoths. It continued in use until the fourteenth,
and in Portugal even until the fifteenth, century.
No reckoning is easier to distinguish from all
others, for the date is always given, not as anno
MCIII., but as Era MCIII. (which is equivalent

ERAS 39
to a.d. 1065), and it is free from the confusion which
is found elsewhere, since the year is invariably taken
to begin with January 1.

The Christian Year
In 525 Dionysius Exiguus, a Scythian monk living
in Rome, undertook to compile an Easter Table
to take the place of that which was in use, but of
which the Cycle would terminate in 531. That
Cycle was calculated from the 153rd year of the
Era of Diocletian, which began in a.d. 285 —
therefore from a.d. 437 — and consisted of five
Lunar Cycles or 95 years. In continuing it Diony
sius says that he preferred not to give the years of
a pagan persecutor, but chose rather to set them
out from the Incarnation of Jesus Christ. Con
sequently he marked the year next following
247 from Diocletian, as the Year of our Lord 532.
He makes no suggestion that he is inventing a
new Era; he simply inserts the year calculated
from one of the several slightly discrepant dates
which were computed to indicate the year of our
Lord's birth, and uses it to start a series of years
to accompany his Easter Table. And it never, in
fact, became an Era until it was employed as a
chronological note in English charters late in the

40 MEDIEVAL RECKONINGS OF TIME
seventh century, and was adopted by Bede in his
" Church History of the English Race."
The Year of Grace, therefore — to use the name
familiar in England — is an English invention.
From England it travelled to the Continent, prob
ably in company with St. Boniface. It was inserted
in documents and annals. Afterwards it was taken
up by the Frankish Kings and Emperors ; and from
their chancery it passed into that of the Popes, for
a few years in the tenth century and normally from
1048. It was the only Era in use in the West out
side Spain, and it had no serious competitor.
Occasionally, no doubt, an historical writer would
for special reasons give his dates from the Passion
(taken as a.d. 33) ; or another might invent a new
Era, of no merit, beginning in 22 B.C., but he would
be careful to specify his date as reckoned secundum
evangelicam veritatem. These are eccentricities
which do not affect the truth of our general state
ment. But if the first Year of Grace was a.d. 1, there
was no common agreement as to the date in that
year at which the Incarnation was assumed to have
taken place. It is probable that it was originally
considered to have been that of the Annunciation,
on March 25, 1 B.C.; but Bede laid down that the

ERAS 41
year began with the Nativity on December 25,
1 B.C. Both these modes of reckoning were long
maintained, but before entering into particulars it
will be well to mention some other dates for be
ginning the year which were transmitted from
earlier systems. The Beginning of the Year
The Roman civil year began on January 1, and
calendars could hardly begin on any other day.
Hence the Golden Number and the Sunday Letter
were reckoned from that date; and the years of the
Spanish Era, which were connected with a calendar,
always began on January 1. But Christian opinion
condemned it on account of its association with
pagan festivities. It was sought to convert the
day by the institution of the Feast of the Circum
cision; but to speak, as some writers do, of the
reckoning from January 1 as the Style of the Cir
cumcision is to invert the order of things. The
revival of the study of Roman law led to the use
of this reckoning in the later middle ages, and it
is found occasionally in Germany from the thir
teenth century ; but elsewhere it is the rarest of all
the modes of dating the beginning of the year. Still
it is curious that everywhere the term New Year's

42 MEDIEVAL RECKONINGS OF TIME
Day means January 1 and no other day, and popular
usages have always been specially attached to it.
The Roman year had originally been reckoned
from March 1. But it was for a different reason
that Christian writers often spoke of March as the
first month. It was because that was the month in
which the Passion and Resurrection were believed
to have occurred, and from which all Easter cal
culations proceed. In connection with this the
Concurrents were changed on March 1. The date
seems also to have been taken as the beginning of
the year by the Franks, possibly in continuation
of an old Teutonic custom which made Spring open
with it. It is found in the edicts of Lombard Kings,
and was used at Benevento as late as the twelfth
century. But in one place only was the 1st of
March always officially observed as the beginning
of the year. This was at Venice, where it was
maintained until the fall of the Republic in 1797. !
The Roman financial year, and from it the
original Indiction, began on September 1. This
was rarely used in the West as a date from which to
reckon the year, though it seems that in England
the year was counted in the tenth and early eleventh
centuries as beginning with the Bedan Indiction
of September 24. The date of September 1 is

ERAS 43
strictly a Byzantine date, usually associated with
the Byzantine annus Mundi, which was calculated
from 5509 B.C. It is scarcely found in the
West, except in those parts, Lower Italy and
Sicily, which continued latest under Byzantine
rule. From these three beginnings, of January 1,
March 1, and September 1, we return to those
which were connected with the Year of Grace.
The reckoning from the presumed date of the
Annunciation — that is, from March 25 preceding
a.d. 1 — prevailed in some parts, especially in the
kingdom of Burgundy and in the North of Italy.
It persisted longest at Pisa, whence it is known
as the calculus Pisanus. The Popes used it occa
sionally between 1088 and 1145, but never (it is
believed) before or after.
The inconvenience of this mode of reckoning was
manifest, and the years of the Incarnation came
to be dated, not from the Annunciation, but from
the Nativity, or Christmas Day, December 25, 1 B.C.;
this continued to be the established beginning of
the year until at least the eleventh century. It
held its ground with the Popes until 1088, with
the Emperors until the second quarter of the
thirteenth century, and with the Anglo-Saxon and

44 MEDIEVAL RECKONINGS OF TIME
Norman Kings of England. The few exceptions
to this rule can be accounted for by special circum
stances. Even after the Popes had settled down
to the Lady Day reckoning for great bulls
(privilegia), they used Christmas for all their
official correspondence; and in England, when the
same change was made, the Benedictine monas
teries continued to begin the year at Christmas as
late as the fourteenth century.
The system by which it was supplanted was that
which we know as the style of Lady Day, March 25,
subsequent to the reckoning from Christmas. It
was thus twelve months behind the calculus
Pisanus. This reckoning may perhaps be traced
back to the end of the tenth century. It took root
at Florence, and is hence called the calculus
Florentinus, to distinguish it from the rival system
which prevailed at Pisa. The wider diffusion of
this mode of reckoning is attributed, at least in
part, to the influence of the Cistercians. We find
it in various districts of France, but mainly as an
ecclesiastical use. When the Popes finally adopted
it, under Alexander III., they did so only for their
most solemn documents. But in England under
the Plantagenets it became the regular system:
whenever we find a date specified as secundum

ERAS 45
consuetudinem ecclesice Anglicance, we know that
the year begins at Lady Day.
A computation not far removed from this in
actual date, but the most illogical and incon
venient that could have been devised, is that from
Easter. The number of the year is reckoned
from the Nativity, but its initial day from the
Resurrection; and as the time of Easter may vary
by more than a month, the length of the year is
never uniform, and if Easter falls early in one
year and late in the next, the same year may
include a good many days of March or April at
both ends. We should notice that the Easter date
was considered to include Easter Eve, and is
sometimes spoken of as a Passione. This is not
the rarely used Year of the Passion which has
been mentioned earlier ; it is only the day that
falls within the season of the Passion,
This unhappy system came into use in the
French Court from the beginning of the thirteenth
century; it is known as the mos Gallicanus. But
it never wholly supplanted the reckoning from
Lady Day in the local custom of many districts of
France, and it never travelled far beyond the
limits of the kingdom except to places which were
closely connected with it through the ruling houses

46 MEDIEVAL RECKONINGS OF TIME
or through trading relations. Thus it became
officially used in Holland and even in Cologne;
but it was known as the stylus curice, and it
did not supersede the popular reckoning from
Christmas. If these varieties of reckoning were at times apt
to mislead people in the Middle Ages, much more
have they confused modern historians and editors.
Nor is the usage of a particular place or chancery
decisive as to the system employed in a given
document or narrative. A clerk from Bologna or
Lucca might introduce the rule to which he was
accustomed; or a royal document might be drawn
up in a monastery which counted dates differently
from the King's chancery. But these divergences
were matters for the writer of documents or
chronicles; the Year of Grace did not practically
concern the common man. If we suppose a
traveller to set out from Venice on March 1, 1245,
the first day of the Venetian year, he would find
himself in 1244 when he reached Florence; and
if after a short stay he went on to Pisa, the year
1246 would have already begun there. Continuing
his journey westward, he would find himself again
in 1245 when he entered Provence, and on arriving
in France before Easter (April 16) he would be

ERAS 47
once more in 1244. This seems a bewildering
tangle of dates. But, in fact, our traveller would
not think of the year; he would note his move
ments by the month and day. If he entered the
year at the beginning of his diary, it would be the
year of the place where he lived, not of the places
through which he passed.

3 9002 03230