In classical Latin even before the time of Christ it was usual for correspondents to indicate when and where their letters were written. This was commonly done by such words as dabam Romanæante diem quartum Kalendas Januarias, i.e. I gave or delivered this at Rome on December 29th. For this the later formula was data Romæ (given at Rome). Hence data, the first word of the formula, came to be used for the time and place therein specified. The principle that imperial decrees and charters must be "dated" as a condition of validity, i.e. that they must bear upon them the indication of the day and year when they were delivered, may be traced back to the time of Constantine. In the course of the Middle Ages this principle was generally admitted, and we find, for example, that at Cologne in the twelfth century the validity of a certain instrument was contested because it lacked a date. "Those who have seen it say that the document which John brought does not bear the day or the indiction . . . now the Roman decrees lay down that letters which lack the day and the indiction have no binding force." (Westdeutsche Zeitschrift für Geschichte, I, 377.) But although this principle was recognized in theory it was not always carried out in practice. Even down to the beginning of the twelfth century not only royal and imperial letters but even charters (Urkunden), properly so called, were occasionally through the carelessness of officials sent out without a date. (Bresslau, Handbuch, I, 891.) In this matter the Italian chancery officials seem to have been much more careful than those of the rest of Europe. The same is true with regard to the correctness of the dates which do appear in official documents, especially those of the early Middle Ages. As a rule the charters emanating from the chancery of the Western Emperors are much more liable to this form of error than those of the Holy See (Bresslau, ib., 844). But even the bulls of such a pontiff as Innocent III are not unfrequently at fault, and as Léopold Delisle has shown, an erroneous calculation of the indiction may be perpetuated through a whole series of authentic documents (Bib. de l'Ecole des chartes, 1858, p. 55). In any case it remains certain and is admitted by all serious writers upon diplomatics that the mere fact that an erroneous date occurs in a document, especially when we are dealing with the earlier Middle Ages, cannot by itself be accepted as a proof, or even a presumption, of the spuriousness of the document.

THE CHRISTIAN ERA

The point of main interest in this connection is to determine the source and period of the introduction of our present system of dating by the Christian Era. Although, as explained in the article GENERAL CHRONOLOGY, the monk known as Dionysius Exiguus, when resident in Rome, c. 527, seems to have been the first to initiate the practice of calculating years from the birth of Christ and although it was undoubtedly he who identified the year of Christ's birth with the year 753 of the foundation of Rome, as is still done in our current chronology, nevertheless it was not until long after the age of Dionysius Exiguus that the system came into common use. For example, no trace of it will be found in that great historian of the Gallic Church, St. Gregory of Tours, the contemporary of our St. Augustine of England; and in the writings of Pope St. Gregory the Great the Dionysian Era is not adopted. It was the pope's habit to date his letters by the regnal years of the emperor and letters so dated may be seen in Bede's "Ecclesiastical History", just as they were copied from the Roman archives. Apparently it was the Englishman Bede himself who was the first to bring the Dionysian system into general use, for it was through him that it was adopted in literature, having been employed systematically not only in his "De Temporum Ratione" but especially in his "Ecclesiastical History". What is more, we may notice the striking fact that the regular employment of the Christian Era in English charters began just at the period of Bede's pre-eminent influence. It is only from about the year 679 that we are able to appeal to English charters of indisputable authenticity. Taking eight such documents, the eight earliest which we can quote with confidence and dated respectively 679, 692, 697, 732, 734, 736, 740, 759, we may notice says Professor Earle (Land Charters, Introduc., p. xxxiii) that "of this series the first five though all more or less dated, whether by the month or the regnal year, or by the Indiction, or by all these at once, have not the Anno Domini. On the other hand, the last three agree in using the Christian Era and from this time the practice is continuous. In the intervening year which breaks this series into two parts falls the death of Bede A.D. 735." Very noteworthy is the decree of an English synod held in 816, wherein it is prescribed that the bishop shall put the acts of the synod into writing and date them by the Era of the Incarnation. This points no doubt to a time "when ecclesiastics knew the era well enough but had not yet acquired the punctual habit of using it". It is in any case certain that neither in the papal chancery nor in that of the Western Empire was the system introduced until considerably later. In the empire it only became general in the latter part of the ninth century, while although it occurs occasionally in papal documents of the time of John XIII (965-972), it was not the rule before the twelfth century. But for the dating of papal documents and for the so-called "double date" see the article BULLS AND BRIEFS.

RECKONING OF YEARS

Before the Christian Era was generally adopted in the dating of documents various other systems were employed at different periods and in different countries. The best known of these was the counting by "indictions". The indiction was a cycle of fifteen years, the first of these cycles being conceived to have started at a point three years before the beginning of the present Christian Era. It was usual to indicate only the position of the year in the current indiction, and no notice was taken of the number of cycles already completed. Thus, for example, indictio quarta meant the fourth year of some particular indiction and not the fourth cycle of fifteen years after B.C. 3; from which it follows that merely to know the year of the indiction is useless for determining the absolute date of any document unless we know otherwise approximately the period to which the document belongs. In reckoning the beginning and consequently the determining-point of the indiction-cycles four different systems were adopted: the indictio Græca according to which the indiction began on September 1st; the indictio Cæsarea (or indiction of Bede) beginning September 24th; the indictio Romana beginning December 25th or January 1st; and the indictio Senensis beginning September 8th. The indictio Græca was the oldest of these and it remained in use in papal bulls until 1087 and in imperial documents until 832. It was partly supplanted, especially in the Carlovingian sphere of influence, by the indictio Cæsarea.

Concurrently with the year of the indiction it was customary both in papal and imperial documents to mention the regnal year of the pope or emperor. So far as regards the emperors this was prescribed by Justinian (Novella xlvii). In the case of the popes we do not know any instance earlier than 787. Generally speaking (though the rule admitted of many exceptions, especially later) the regnal year was calculated from the day of coronation or consecration. In the official acts of most of the countries of Christendom, and notably in England, the regnal year of the sovereign was always given and sometimes this was the only indication of the year. As a continuous system of year enumeration the oldest era in practical use appears to have been that known as the "Era of the Martyrs" or "of Diocletian" (anni Diocletiani). Its starting-point was the accession of the Emperor Diocletian, 29 August, 284. The Spanish Era (æra Hispanica) was in familiar use in Spain from the fifth century down to late in the Middle Ages. It adds about thirty-eight years to the ordinary numbering of the Christian Era. Where Byzantine influences prevailed the years were generally numbered from the beginning of the world (ab origine mundi). This era was calculated from 1 September, and the birth of Christ, which is the point of departure of our present chronology, took place in the year 5509 of the Byzantine system. Several other methods of reckoning, of which the best known is the Era of the Hegira followed in Mahommedan countries, have also prevailed in various localities, but they cannot be discussed in detail here. After the Christian Era had been universally adopted an important source of confusion as regards the dating of documents still remained in the diversity of practice about the beginning of the year. For the details of this the reader must be referred to the article GENERAL CHRONOLOGY, but we may notice here that among the Anglo-Saxons, as also at many different periods in the papal chancery, the new year was calculated to begin on December 25th. On the other hand, in England from the twelfth century onwards, largely under Norman influences, the years were numbered from the 25th of March. This arrangement was often called the mos Anglicanus or computatio Anglicana, though it also prevailed in Florence, Siena, Pisa, and at least occasionally in other parts of the Continent as well as in many papal documents. In England it lasted on down to the eighteenth century, though after Elizabethan times it became increasingly common in the dating of letters to indicate the system of dating adopted, N.S. often standing for the New or Continental Style in which the year began on January 1st, and 0.S. for the Old Style in which the year was counted from March 25th. Further N.S. was still more frequently used for dates which followed the reformed calendar of Gregory XIII, as explained in the article GENERAL CHRONOLOGY.

The Reckoning of Days

The early converts to Christianity in the West not unnaturally retained the method of indicating the days of the month which was current among their pagan contemporaries. According to this, three fixed points were taken in each month, the kalends on the first day, the ides on the thirteenth (or in some months on the fifteenth), and the nones on the ninth day before the ides and consequently on the fifth or seventh. The dates which fell between these fixed points were designated by the number of days by which they fell short of the next fixed point. Thus the twenty-fourth of May was called ante diem nonum kalendas Junias (i.e. the ninth day before the kalends of June). During the early Middle Ages this system was retained practically unaltered except that the long Roman form was somewhat contracted, for example decimo kalendas Julii was written instead of the ante diem decimum kalendas Julias.

A curious arrangement prevailed at Bologna (it was called from its place of origin the consuetudo Bononiensis) and extended over a large part of the north of Italy. According to this the first half of the month was numbered forwards and called menses intrans, but the last half of the month was called mensis exiens and numbered backwards, as in the Roman system; thus the seventeenth of May was called die quinto decimo exeuntis mensis Maii. Our present system of numbering the days straight on from the first of each month began to appear in the sixth century and gradually became more prevalent throughout the Middle Ages, but it never came into general use on account of the custom of indicating the day by the feasts of the local calendar. Not only did the common festivals serve for this purpose, but the Sundays were also often used, and were designated by the first words of their Introit in the Mass. For instance in Dominica lætare means on the Sunday whose Introit begins with Lætare in Domino, i.e. the fourth Sunday of Lent. Moreover the vigil of a feast, or the previous day, or the octave, or a specified day within the octave, were all familiarly designated by their relation to the feast, e.g. in pervigilio Nativitatis Beatæ Mariæ; postridie Sancti Laurentii; in octava Sti. Laurentii, etc. In this method of dating, which was constantly employed both in Latin and ill the vernacular, the use of the English word utas for octave should be noticed. This method of dating by saints' days, as will be readily understood, depended much upon local conditions and was always apt to become both complicated and inexact.

HERBERT THURSTON