We do not know how soon an effort was made to combine in one book the several portrayals of the life of Jesus. Luke in his Gospel (1:1-4) makes a selection of the material and incorporates data from different sources, but with the stamp of his own arrangement and style. He followed, in the main, the order of Mark's Gospel, as is easily seen. But this method is not what is meant by a harmony of the Gospels, for the result is a selection from all sorts of material (oral and written), monographs and longer treatises.

The first known harmony is Tatian's Diatessaron (dia tessaron, by four) in the second century (about 160 A.D.) in the Syriac tongue. It was long lost, but an Arabic translation has been found and an English rendering appeared in 1894 by J. Hamlyn Hill. It is plain that Tatian has blended into one narrative our Four Gospels with a certain amount of freedom as is shown by Hobson's The Diatessaron of Tatian and the Synoptic Problem (1904). There have been modern attempts also to combine into one story the records of the Four Gospels. There is a superficial advantage in such an effort in the freedom from variations in the accounts, but the loss is too great for such an arbitrary gain. The word harmony calls for such an arrangement, but it is not the method of the best modern harmonies which preserve the differences in material and style just as they are in the Four Gospels.

In the third century Ammonius arranged the Gospels in four parallel columns (the Sections of Ammonius). This was an attempt to give a conspectus of the material in the Gospels side by side. In the fourth century Eusebius with his Canons and Sections enabled the reader to see at a glance the parallel passages in the Gospels. The ancients took a keen interest in this form of study of the Gospels, as Augustine shows.

Of modern harmonies that by Edward Robinson has had the most influence. The edition in English appeared in 1845, that in Greek in 1846. Riddle revised Robinson's Harmony in 1889. There were many others that employed the Authorized Version, like Clark's, and that divided the life of Christ according to the feasts.

Broadus (June, 1893) followed Waddy (1887) in the use of the Canterbury Revision, but was the first to break away from the division by feasts and to show the historical development in the life of Jesus. Stevens and Burton followed (December, 1893) Broadus within six months and, like him, used the Canterbury Revision and had an independent division of the life of Christ to show the historical unfolding of the events. These two harmonies have held the field for nearly thirty years for students of the English Gospels. In 1903 Kerr issued one in the American Standard Version and James one in the Canterbury Revision (1901).

Harmonies of the Gospels in the Greek continued to appear, like Tischendorf's (1851, new edition 1891), Wright's A Synopsis of the Gospels in Greek (1903), Huck's Synopse der drei ersten Evangelien (1892, English translation in 1907), Campbell's First Three Gospels in Greek (1899), A Harmony of the Synoptic Gospels in Greek by Burton and Goodspeed (1920).

The progress in synoptic criticism emphasized the difference in subject matter and style between the Synoptic Gospels and the Fourth Gospel as appears in the works of Huck, Campbell, and Burton and Goodspeed that give only the Synoptic Gospels. Burton and Goodspeed have also an English work, A Harmony of the Synoptic Gospels for Historical and Critical Study (1917). In 1917 Sharman (Records of the Life of Jesus) gives first a harmony of the Synoptic Gospels with references to the Fourth Gospel and then an outline of the Fourth Gospel with references to the Synoptic Gospels.

Once more in 1919 Van Kirk produced The Source Book of the Life of Christ which is only a partial harmony, for the parables and speeches of Jesus are only referred to, not quoted. But he endeavored to show the results of Gospel criticism in the text of the book. There is much useful material here for a harmony, but it is not a real harmony that can be used for the full story of the life of Jesus. Van Kirk, however, is the first writer to place Mark in the first column instead of Matthew. I had already done it in my outline before I saw Van Kirk's book, but his was published first. It is an immense improvement to put Mark first. The student thus sees that the arrangement of the material is not arbitrary and whimsical, but orderly and natural. Both Matthew and Luke follow Mark's order except in the first part of Matthew where he is topical in the main. John supplements the Synoptic Gospels, particularly in the Judean (Jerusalem) Ministry.

Slowly, therefore, progress has been made in the harmonies of the Gospels. But the modern student is able to reproduce the life and words of Jesus as has not been possible since the first century. It is a fourfold portrait of Christ that we get, but the whole is infinitely richer than the picture given by any one of the Four Gospels. The present Harmony aims to put the student in touch with the results of modern scholarly research and to focus attention on the actual story in the Gospels themselves. One may have his own opinion of the Fourth Gospel, but it is needed in a harmony for completeness.

The criticism of the synoptic gospels has been able to reach a broad general conclusion that is likely to stand the test of time. The reason for this happy solution lies in the fact that the processes and results can be tested. It is not mere subjective speculation. Any one who knows how to weigh evidence can compare Mark, Matthew, and Luke in the English, and still better in the Greek. The pages of the present harmony offer proof enough. It is plain as a pikestaff that both our Matthew and Luke used practically all of Mark and followed his general order of events. For this reason Mark has been placed first on the pages where this Gospel appears at all. But another thing is equally clear and that is that both Matthew and Luke had another source in common because they each give practically identical matter for much that is not in Mark at all. This second common source for Matthew and Luke has been called Logia because it is chiefly discourses. It is sometimes referred to as "Q", the first letter of the German word Quelle (source). Unfortunately we do not have the whole of the Logia (Q) before us as in the case of Mark, though we probably do not possess the original ending of Mark in 16:9-20. But we can at least reproduce what is preserved. Still, just as sometimes either Matthew or Luke made use of Mark, so in the case of the Logia that is probably true. Hence we cannot tell the precise limits of the Logia. Besides, a small part of Mark is not employed by either Matthew or Luke and that may be true of the Logia. But the fact of these two sources for Matthew and Luke seems to be proven.

But there are various other points to be observed. One is that both Matthew and Luke may have had various other sources. Luke tells us (Luke 1:1-4) that he made use of "many" such sources, both oral and written. And a large part of Luke does not appear in the other gospels or at least similar events and sayings occur in different environments and times. Hence our solid conclusion must allow freedom and flexibility to the writers in various ways. We can see for ourselves how Matthew and Luke handled both Mark and the Logia, each in his own way and with individual touches of style and purpose.

One other matter calls for attention. Papias is quoted by Eusebius as saying that Matthew wrote in Hebrew (or Aramaic) whereas our present Matthew is in Greek. It is now commonly held that the real Matthew (Levi) wrote the Logia first in Aramaic and that either he or some one else used that with Mark and other sources for our present Gospel of Matthew.

It should be added also that there is a considerable body of evidence for the view that Mark wrote under the influence of Simon Peter and preserves the vividness and freshness of Peter's own style as an eyewitness.

One other result has come. It is increasingly admitted that the Logia was very early, before 50 A.D., and Mark likewise if Luke wrote the Acts while Paul was still alive. Luke's Gospel comes (Acts 1:1) before the Acts. The date of Acts is still in dispute, but the early date (about A.D. 63) is gaining support constantly. The upshot of these centuries of synoptic criticism has brought into sharp outline the facts that now stand out with reasonable clearness. There are many points in dispute still, but we at least know how the synoptic gospels were written, and are reasonably certain of the dates and the authors.

There are many good books on the subject, like Hawkin's Horae Synopticae (second edition), Sanday's Oxford Studies in the Synoptic Problem, Harnack's Sayings of Jesus and his Date of the Synoptic Gospels and Acts. My own views appear in my Commentary on Matthew (Bible for Home and School), Studies in Mark's Gospel, and Luke the Historian in the Light of Research.

But a considerable body of scholars still hold that the same man wrote both the Gospel and the Apocalypse, but a different explanation is offered by two groups. One class of writers affirm that John wrote the Apocalypse first before he had come to be at home in the Greek idiom as we see it in the Gospel and the Epistles. We know that John and Peter were fishermen and were not considered men of literary training by the Sanhedrin (Acts 4:14). This explanation is sufficient but for the further fact that the early date of the Apocalypse (about 70 A.D.) is not now so generally held to be true. The later or Domitianic date as given by Irenæus seems pretty clearly to be correct. So the other group suggest that the books may belong substantially to the same period (the Domitianic date) and that the explanation of the grammatical infelicities in the Apocalypse may be due to the fact that John being on the Isle of Patmos when he wrote did not have the benefit of friends in Ephesus who apparently read the Gospel (John 21:24-25). Besides, the excited state of John's mind because of the visions may have added to the number of the solecisms in the Apocalypse. This view I personally hold as probable. The unity of both Gospel and Apocalypse is denied by some.

So the matter stands as between the Gospel and the Apocalypse. But the Fourth Gospel has difficulties of its own. These relate in part to the book in itself. It is true there is a great similarity in language and style between the narrative parts of the book and the discourses of Jesus. It is affirmed that the writer has colored the speeches of Jesus with his own style or even made up the dialogues so that they are without historical value or at least on a much lower plane than the Synoptic Gospels as objective history. There is something in this point, but one must remember that the Synoptic Gospels vary in their manner of reporting the speeches of Jesus and aim to give the substance rather than the precise words of the Master in all instances. It is at most a matter of degree. There is a Johannine type of thought and phrase beyond a doubt, but curiously enough we have a paragraph in Matthew 11:24-31 and Luke 10:21-23 that is precisely like the Johannine specimens, written long before the Fourth Gospel. One must remember the versatility of Jesus, who could not be retained in any one style or mold. But there are those who admit the Johannine authorship of the Gospel and yet who refuse to put it on the same plane as the Synoptic Gospels. Every one must decide for himself on this point. For myself I see too much of Christ in the Fourth Gospel in the most realistic and dramatic form to be mere invention. We can enlarge our conception of Christ to make room for the Fourth Gospel.

But even so it is urged that the Beloved Disciple cannot be the Apostle John. If not, then the Fourth Gospel ignores the Apostle John,—a very curious situation. It is a long story for which one must go to the able books in defense of the Johannine authorship by Ezra Abbott, James Drummond, W. Sanday, Luthardt, Watkins and many others. The ablest modern attacks are made by Bacon and Wendt and Schmiedel. My own view is given in my The Divinity of Christ in the Gospel of John.

(a) It is objected that Matthew is mistaken in making three sets of fourteen each. There are only forty-one names, and this would leave one set with only thirteen. But does Matthew say he has mentioned forty-two names? He does say (1:17) that there are three sets of fourteen and divides them for us himself: "So all the generations from Abraham unto David are fourteen generations; and from David unto the carrying away to Babylon fourteen generations; and from the carrying away to Babylon unto the Christ fourteen generations." The points of division are David and the captivity; in the one case a man, in the other an event. He counts David in each of the first two sets, although Jechoniah is counted only once. David was the connecting link between the patriarchal line and the royal line. But he does not say "from David to Jechoniah," but "from David to the carrying away unto Babylon," and Josiah is the last name he counts before that event. And so the first name after this same event is Jechoniah. Thus Matthew deliberately counts David in two places to give symmetry to the division, which made an easy help to the memory.

(b) The omissions in Matthew's list have occasioned some trouble. These omissions are after Joram, the names of Ahaziah, Joash, Amaziah, and after Josiah, these of Jehoiakim and Eliakim (2 Kings 8:24; 1 Chron. 3:11; 2 Chron. 22:1, 11; 24:27; 2 Kings 23:34; 24:6). But such omissions were very common in the Old Testament genealogies. See 2 Chron. 22:9. Here "son of Jehoshaphat" means "grandson of Jehoshaphat." So in Matt. 1:1 Jesus is called the son of David, the son of Abraham. A direct line of descent is all that it is designed to express. This is all that the term "begat" necessarily means here. It is a real descent. Whatever omissions were made for various reasons, would not invalidate the line. The fact that Ahaziah, Joash, and Amaziah were the sons of Ahab and Jezebel would be sufficient ground for omitting them.

(c) Matthew mentions four women in his list, which is contrary to Jewish custom, viz. Tamar, Rahab, Ruth, and the wife of Uriah. But neither one is counted in the lists of fourteen, and each one has something remarkable in her case (Broadus, Comm. on Matt. in loco). Three were guilty of gross sin, and one, Ruth, was of Gentile origin and deserved mention for that reason. This circumstance would seem to indicate that Matthew did not simply copy the genealogical history of Joseph. He did this, omitting what suited his purpose and adding likewise remarks of his own. His record is thus reliable and yet made a part of his own story.

2. A comparison of the lists of Matthew and Luke.

If no list had been given by Luke, no further explanations would be necessary. But Luke not only gives a list, but one radically different from Matthew's, and in inverse order. Matthew begins with Abraham and comes to Jesus; Luke begins with Jesus and concludes with Adam [the son of God]. Several explanations are offered to remove the apparent contradiction.

(a) As early as Julius Africanus it was suggested that the two lines had united in accordance with the law of Levirate marriage. By this theory, Heli and Jacob being stepbrothers, Jacob married Heli's widow and was the real father of Joseph. Thus both genealogies would be the descent of Joseph, one the real, the other the legal. This theory is ably advocated by McClellan, pp. 416 ff., and Waddy, p. xvii. It is argued that Jechoniah's children were born in captivity and so, being slaves, lost both his royal dignity and his legal status. Stress is laid upon the word "begat" to show that Matthew's descent must be the natural pedigree of Joseph, and upon the use of the expression "son (as was supposed) of Joseph." Hence both Joseph's real and legal standing are shown, for by Luke's account he had an undisputed legal title to descend from David. This is certainly possible, although it rests on the hypothesis of the Levirate marriage.

(b) Lord Arthur Hervey, in his volume on the Genealogies of Our Lord, and in Smith's Dictionary, argues that Matthew gives Joseph's legal descent as successor to the throne of David. According to this theory Solomon's line failed in Jechoniah (Jer. 22:30) and Shealtiel of Matthew's line took his place. Luke's account, on the other hand, gives Joseph's real parentage. Matthew's Matthan and Luke's Mattathias are identified as one, and the law of Levirate marriage comes into service with Jacob and Heli. This explanation has received favor with such writers as Mill, Alford, Wordsworth, Ellicott, Westcott, Fairbairn. McNeile (on Matthew) considers this the "only possible" view. The chief objection seems to be the most natural meaning of "begat," implying direct descent, and the necessity for two suppositions, one about Shealtiel and another about Jacob and Heli. It is even fairly probable that the Shealtiel and Zerubbabel of Matthew and Luke are different persons.

(c) The third and most plausible solution yet suggested makes Matthew give the real descent of Joseph, and Luke the real descent of Mary. Several arguments of more or less weight can be adduced for this hypothesis.

(1) The most natural meaning of "begat" in Matthew is preserved. Jesus goes through David's royal line and so fulfils prophecy. It is not elsewhere stated that Mary was of Davidic descent, although presumptive evidence exists in the language of the angel (Luke 1:32) and the enrollment of Mary (Luke 2:5). So Robinson (Revised edition).

(2) The use of Joseph without the article, while it is used with every other name in the list. "The absence of the article puts the name outside of the genealogical series properly so-called."—Godet. This would seem to indicate that Joseph belonged to the parenthesis, "as was supposed." It would read thus, "being son (as was supposed of Joseph) of Heli." Luke had already clearly stated the manner of Christ's birth, so that no one would think he was the son of Joseph. Jesus would thus be Heli's grandson, an allowable meaning of "son." See Andrews' (new edition) Life of Our Lord, p. 63.

(3) It would seem proper that Matthew should give the legal descent of Jesus, since he wrote chiefly for Jews. This, of course, could only be through Joseph.

(4) And it would seem equally fitting that Luke should give the real genealogy of Jesus, since he was writing for all. And this could come only through Mary. If it is objected that a woman's genealogy is never given, it may be replied that women are mentioned for special reasons in Matthew's list, though not counted, and that Mary's name is not mentioned in this list. The genealogy goes back to her father either by skipping her as suggested above and making son mean the grandson of Heli, or by allowing Joseph to stand in her place in the list, as he would have to do anyhow. On the whole, then, this theory seems the most plausible and pleasing. So practically Luther, Bengel, Olshausen, Lightfoot, Wieseler, Robinson, Alexander, Godet, Weiss, Andrews (new edition, p. 65), Broadus, and many recent writers.

But Bacon (Genealogy of Jesus Christ, Hastings D. B. and Am. J. of Theol. Jan., 1911) says that nearly all writers of authority abandon any effort to reconcile the two pedigrees of Jesus save as the effort of Christians to give "His Davidic sonship rather than His actual descent." See Machen's survey of negative criticism, on the subject in Princeton Theol. Review (Jan., 1906). Barnard (Hastings D. C. G.) admits two independent accounts, but sees no solution, but Sweet (Int. St. Bible Encyl.) accepts the view that Matthew gives the real genealogy of Joseph and Luke that of Mary. Plummer (Comm. on Luke) thinks it incredible that Mary's genealogy should be given by Luke.

1. The death of Herod the Great is relied on with most certainty to fix the year of Christ's birth. The rule of Archelaus and Antipas demands B.C. 4. Josephus mentions an eclipse of the moon which occurred shortly before he died. Ant. XVII, 6, 4. This eclipse is the only one alluded to by Josephus, and fixes with absolute certainty the time after which the birth of Jesus could not have occurred, since, according to Matt. 2:1-6, Jesus was born while Herod was still living. The question to be determined would be the year of this eclipse. Astronomical calculations name an eclipse of the moon March 12 and 13, in the year of Rome 750, and no eclipse occurred the following year that was visible in Palestine. Josephus (Ant. XVII, 8, 1), says that Herod died thirty-seven years after he was declared king by the Romans. In 714 he was proclaimed king, and this would bring his death counting from Nisan to Nisan, as Josephus usually does, "in the year from 1st Nisan 750 to 1st Nisan 751, according to Jewish computation, at the age of seventy" (Andrews). Herod died shortly before the Passover of 750, then, according to the eclipse and the length of his reign. Caspari contends for January 24, 753, as the date of Herod's death, because there was a total eclipse of the moon January 10. So he puts his death fourteen days later. Mr. Page (New Light from Old Eclipses) argues for the eclipse that occurred July 17, 752, as the one preceding Herod's death. He thinks that this makes unnecessary the subtraction of two years from the reign of Tiberius on the theory that Tiberius was contemporary ruler with Augustus for two years. But he finds difficulty in lengthening Herod's reign so long, and his theory has gained no great acceptance as yet. Our present era makes the birth of Christ in the year of Rome 754, and is due to the Abbot Dionysius Exiguus in the Sixth Century. Hence it is clear that if Herod died in the early spring of 750, Jesus must have been born at least four years before 754, the common era, and likely in the year 749.

2. It has been inferred by some that Jesus was at least two or three years old when Herod slaughtered the infants in Bethlehem, Matt. 2:16. Thus the year would be put two years further back to the end of 747 or beginning of 748. But this is not demanded by the "two years" of Matthew, for Herod would naturally extend the limit so as to be sure to include the child in the number slain, and a child just entering the second year would be called "two years" old by Jewish custom. No more definite note of time comes from this circumstance, save that the massacre probably took place some months before Herod's death, which fact would bring the Saviour's birth back some time into the year 749.

3. The appearance of the "star in the east" (Matt. 2:2). This, of course, was before Herod's death, and would agree in time with the slaughter of the children, if the star be looked upon as a supernatural phenomenon, and not the wise men's interpretation of a natural conjunction of planets. Kepler first suggested that, as there was a conjunction of Jupiter and Saturn in 747, to which Mars was added in 748, this conjunction might have been the bright star that led on the wise men. See Wieseler, Synopsis, p. 57. Kepler had also suggested that a periodical star or a comet might have joined the constellation. The Chinese records preserve the account of the appearance of a comet in the spring of 749. Either of these theories is fascinating in itself, especially to those minds that prefer a natural explanation of anything that looks miraculous. Both phenomena are possible in themselves, but they hardly meet the requirements of the record in Matthew. (1) The word used is aster, star, and not astron, a group of stars. (2) Rev. C. Pritchard, whose calculations have been verified at Greenwich (Smith's Dic.), has shown that those "planets could never have appeared as one star, for they never approached each other within double the apparent diameter of the moon." So Ideler's hypothesis that the wise men all had weak eyes seems rather feeble. (3) The year 747 would conflict slightly with other evidence for Christ's birth that favors 749, although Wieseler, p. 53, note 4, contends that the star first appeared to the wise men two years before their visit, and a second time on their visit to Bethlehem. (4) Besides, the star is said to have stood over "where the young child was," v. 9. If it were a natural star it would have kept going as they went, and would not have stopped till they stopped. Even then it would appear as far away as ever from Bethlehem. It seems best, therefore, to admit the existence of a miracle here, and hence gain nothing from the visit of the Magi to establish the date of the Saviour's birth, save that it was not long before the slaughter of the infants, and would at least agree with the date 749. See Broadus, Comm. in loco.

4. The language of the heavenly host in Luke 2:14 is urged by some as fixing the birth at a time when there was universal peace throughout the world. The closing of the temple of Janus in the time of Augustus is also adduced, but it is not certainly known when it was closed apart from 725 and 729. It was intended to be closed at the end of 744, but was delayed on account of trouble among the Daci and Dalmatæ. See Greswell i. 469. Nothing specific can be obtained from this fact, save that there was a time of comparative quiet in the Roman world from 746 to 752. There was a hush in the clangor of war when Jesus was born.

5. The entrance of John the Baptist upon his ministry gives us another note of time. See Luke 3:1 f. John emerged from the wilderness seclusion in the fifteenth year of the reign of Tiberius. Augustus died August 29, 767. Adding fifteen years to this, the fifteenth year of Tiberius would begin August 29, 781. John was of a priestly family and so could naturally enter upon his work when thirty years of age. Thirty years subtracted from this gives 751, as the date of John's birth. But that is too late by two years to agree with the other date. Here, however, the Roman histories come to our help. Tacitus, Ann. 1, 3: "Tiberius is adopted by Augustus as his son, and colleague in empire." Vell. Pat. 2, 121; "At the request of Augustus, Tiberius was invested with equal authority in all the provinces." So Suetonius Aug. 97 and Tib. 21. It is clear, then, that Tiberius reigned jointly with Augustus about two years before he assumed full control of the empire at the death of Augustus. Luke could have used either date, but Tiberius' power was already equal to that of Augustus in the provinces two years before his death. Luke would naturally use the provincial point of view. Taking off the two years from the joint reign of Augustus, we again come to the year 749, as John was born six months before Jesus. So if John was born in the early part of the spring, Jesus would have been born in the summer or fall of 749.

6. The age of Jesus at his entrance upon his ministry, Luke 3:23. "And Jesus himself, when he began to teach, was about thirty years of age." So most modern scholars, taking the language in the obvious sense. Origen refers it to the beginning of a new life, by the second birth of baptism, after his spiritualizing fashion. The Authorized Version has it: "And Jesus himself began to be about thirty years of age," applying the "beginning" to the period of thirty years. McClellan argues that it means "about thirty years, beginning"; that is, a little the rise of thirty years. The Revised Version seems to be preferable and the only doubt would be as to what is included in the phrase "about thirty years." It has been variously argued that Jesus was from one to three years younger or older than thirty. It seems more reasonable to give the words the meaning that he was just about thirty, a few months under or over. Apparently this fact explains the idiom. The argument that Jesus had to be exactly thirty years old because the priest had to be so, when he entered upon his work, has no great force. For Jesus was not a priest save in a spiritual sense. John had been preaching no great while when Jesus was baptized by him and so entered upon his public ministry. If John began his ministry when he was thirty years old in the fifteenth year of Tiberius, then Jesus's ministry would begin about six months later. His birth would then come in the latter part of 749, unless John was born in the latter part of 748, when it would be earlier in the year.

7. The building of the temple of Herod gives a further clue to the date of Christ's birth. In John 2:20, the Jews say, "Forty and six years was this temple in building." Josephus tells us in one place that Herod began rebuilding the temple in the fifteenth year of his reign, War. I, 21, 1, and in another that he did so in the eighteenth year of his reign, Ant. XV, 11, 1. In the account of Herod's death, Ant. XVII, 8, 1, he used two dates for his reign, according as he counted from his declaration as king by the Romans 714, or the death of Antigonus 717. Eighteen and fifteen would both be correct, according as he reckoned from the one date or the other. Eighteen added to forty-six and both to 714 would make 778. It was at the first Passover in his ministry that this expression is used. It has been probably six months since his baptism. If thirty and a half years be taken from 778, his birth would be thrown back to the year 747, unless the forty-six years be taken as completed, when it would be 748. So Robinson. But this does not quite agree with the other notes of time we have. Many modern harmonists count the eighteen years from 717, and so bring the whole number, adding forty-six, down to 780, or, if the years are complete, 781. Thirty and a half from this would give the autumn of 749 or 750. This is done because Josephus usually reckons Herod's reign from the death of Antigonus, 717. On the whole it seems clear that Josephus is wrong in the War. It is common enough to find Josephus in one passage contradicting what he has said elsewhere. The temple was begun the year that the Emperor came to Syria, as is plain from Josephus. According to Dio Cassius, LIV, 7, this visit was made in B.C. 20 or 19. Correcting Josephus by himself and by Dio Cassius we thus again get B.C. 5 as the probable year of the birth of Christ. See Schuerer, History of the Jewish People in the Time of Jesus Christ, Div. I., Vol. I., p. 410.

8. The census of Augustus Cæsar mentioned in Luke 2:1 f., furnishes the last note of time for this event. This subject is involved in a great many difficulties, and for a full discussion, the reader is referred to Ramsay's Was Christ Born at Bethlehem, and his Bearing of Recent Discovery on the Trustworthiness of the New Testament (Chap. XX) and to my Luke the Historian in the Light of Research. Every statement made by Luke in 2:1-7 was once challenged. Every one is now shown to be correct.

(1) It used to be said that no census was ever taken by Augustus, but heathen writers mention three, in 726, 746, 767. One of these, 746, may be the one here mentioned, which was delayed for various reasons, or which was executed slowly in the distant provinces. But it is not necessary that the phrase "all the world" should be pressed to its literal meaning, though this is more natural. Nor does the argument from silence prove that no other general census was taken by Augustus. But Ramsay has triumphantly vindicated Luke and the general census under Augustus by proof from the papyri that Augustus inaugurated a periodical census every fourteen years from B.C. 8 on. The second occurred A.D. 6 (Acts 5:37). See Ramsay's Was Christ Born at Bethlehem, and Bearing of Recent Discovery on Trustworthiness of the New Testament (Chap. XX) and my Luke the Historian (Chap. XX). We have only to think that there was delay in the carrying out of the census in Palestine to bring this date down to B.C. 6 (or even 5).

(2) It is not a "taxing," but an "enrollment" (Rev. Ver.) that was taken. There was a taxing later (Acts 5:37). And if it were done while Herod was king, Augustus could not have taxed Judea without Herod's consent. But Herod was not now in good form with Augustus.

(3) This helps to explain another objection that the enrollment would not have included Judea anyhow, because it was not yet a province, but a kingdom. But it is not likely that Herod would have displeased Augustus by refusing such information if it was desired. Tacitus asserts that the regna, the dependent kingdoms, were included in the census taken by Augustus.

(4) Hence, also, it is natural that the enrollment should have taken place according to the Jewish and not according to the usual Roman method, because Herod would wish it to be in accordance with the customs of his kingdom. So every one went to his own city. We now know from numerous papyri that in Egypt the family went to the home city. The Jews were used to enrollment by tribes and that was allowed. See Deissmann's Light from the Ancient East, p. 268, and Ramsay's Was Christ Born at Bethlehem, p. 108.

(5) We now have to meet the objection that Quirinius was not governor till ten years later, A.D. 6, when a taxing did occur. (See Acts 5:37.)

It is now possible to give a real solution of this problem. Luke is now shown to be wholly correct in his statement that Quirinius was twice governor, and that the first census took place during the first period. A series of inscriptions in Asia Minor show that Quirinius was governor of Syria B.C. 10-7 and so twice governor of Syria (second time A.D. 6; Josephus, Ant. XVIII, 1:1). See Ramsay, Bearing of Recent Discovery, pp. 273-300, and my Luke the Historian, pp. 127-9. Tertullian (adv. marc. iv, 19) says that Sentius Saturninus was governor of Syria B.C. 9-6. But we now know that Varus was controlling the internal affairs of Syria while Quirinius was leader of the army. Luke is therefore quite accurate in his statement about Quirinius being twice governor of Syria. The Lapis Tiburtinus has iterum Syriam about Quirinius. Ramsay has cleared up this famous historical puzzle and has completely vindicated Luke.

Few subjects have excited as much interest, even needless curiosity, as the date of the birth of the Saviour. But it is noticeable that by the masses of Christians more interest is taken in the day of Christ's birth than in the year. The Christmas festivities and the natural desire to make that the birthday of Jesus cause this widespread interest in December 25. Not only is it impossible to determine with any degree of certainty the day of the month, but the time of the year also is equally uncertain. The chief thing that appears proved is that December 25 is not the time, since the shepherds would hardly be in the fields at night with the flocks, which were usually taken into the folds in November and kept in till March. The nights of December would scarcely allow watching in the mountain fields even as far south as Bethlehem. And besides, the long journey from Nazareth to Bethlehem would hardly be made by Joseph and Mary in winter, the rainy season. McClellan argues for December 25, but his arguments are not convincing. The ancients had various days for Christ's birth: May 20 (Clement of Alexandria), April 20, December 25, January 5. Tertullian and others even say that the day of his birth (December 25) was kept in the register at Rome. But chronologists attach little weight to this testimony, since the same tradition puts the birth of John, June 24; the annunciation of Mary, March 25, and Elizabeth's conception, September 25—the four cardinal points of the year. If one might hazard an opinion, it would be that the birth of Jesus occurred in the summer or early in the fall of 749 or of 748, that is B.C. 6 or 5. Turner (Chronology, Hastings D. B.) reaches B.C. 6 as the probable year of the birth of Jesus though he did not have the new light on the census and on Quirinius which confirms it. Hitchcock (Hastings D. C. G.) saw the bearing of the periodical census that called for B.C. 7-5, but did not yet know the discovery about Quirinius. Armstrong (Chronology New Testament, Int. St. Bible Encycl.) is less certain about the precise year.