Divinity Library

SlSiTL

•YiMLE-VMVEKSJnnf-

DWINITY SCHOOL
TROWBRIDGE LIBRARY

Reprinted from the Journal of Biblical Literature
Vol. XXVIII, Pakt II, 1909

Notes on Gospel Chronology

BENJ. W. BACON
YALE UNIVERSITY

THE New Testament has but one author who writes in the
style and with the purpose of a historian, and but one
date even in this author fully reckoned out according to pre
vailing methods by synchronisms. It is apparent, therefore,
that the date so elaborately fixed in Lk. 3 l is to the author the
all-important date; and this it is, not of course because of
the appearance of John the Baptist, a subordinate figure, but
because it determines the baptism of Jesus and therewith
the beginning of the drama of redemption. The Lucan reck
oning becomes, therefore, the necessary starting point for
modern discussion, as it would seem to have been for patristic
chronography. To its own detriment Gospel chronology has heretofore
been made dependent on harmonization, a method now con
demned on principles of historical criticism. The first postu
late was: All the sources must rest upon the same substantial
datings. In the present discussion the aim will be to ascer
tain and appraise independently all systems of dating pre
sented or implied in the sources considered separately, to
note the extent of agreement, and to draw conclusions upon
comparison of their differences.
A. The Lucan System
1. The Baptism. We have six synchronisms in Lk. 3 l,
" the fifteenth year of Tiberius " and the administrations re
spectively of Pilate (deposed early in 36), Antipas (deposed
in 38), Philip (pb. 33-34), Lysanias (unknown), and Annas
and Caiaphas. Of these the last is incorrectly stated, only

bacon: notes on gospel chronology 131
Caiaphas having been high priest a.d. 18-35. All after the
first are indeterminate. Tib. xv, if the practice of writers
later than Nerva (a.d. 95) was followed, might mean the year
beginning Jan. 1, A.D. 28. More probably the years were
counted as beginning with the accession, Aug. 19, A.D. 14,
in which case Tib. xv would mean the year ending Aug. 19,
A.D. 29. More probably still Luke reckons, like Josephus, from
Passover (e/. Ac. 27 9); i.e. Tib. xv = Apr. A.D. 28-Apr.
A.D. 29. Reckonings from Tiberius' co-regency (ca. 11-12
A.D.) are scarcely admissible. The Baptism, accordingly,
is dated by Luke in the twelvemonth beginning somewhere
from Apr. a.d. 28 to Aug. a.d. 29. The crucifixion might
be in either 29 or 30. Patristic chronographers date it ac
cordingly sometimes as Tib. xv, sometimes Tib. xvi. In
either case they probably mean the same, viz. the " year of
the two Gemini," i.e. of* the consuls L. Rubellius Geminus
and C. Fufius (yar. Rufius, Rufus, Furius) Geminus =
a.d. 29. For the year of the crucifixion was the starting
point in all early reckonings, not that of the beginning of
the ministry, still less that of the nativity. " The year of
the two Gemini" as that of Jesus' death and resurrection
was a constant of primitive Christian chronology which
Turner gives good ground for regarding as having antedated
even our third gospel itself.1
2. The Nativity. Lk. 3 23 gives us a relative dating for
Jesus' birth ; for this is said to have taken place " about "
30 years before the baptism, i.e. B.C. 2-1. This is not seri
ously out of harmony with 1 5, where the opening scenes
of the gospel story— not merely of the birth of the Baptist
— are laid " in the days of Herod the king " (ob. B.C. 4 ; see
1 S.v. " Chronology of the N. T." in Hastings' BD, i. pp. 403 ff. On p.
414 6 Turner rightly emphasizes the fact that the primary chronological in
terest was in the date of the Passion and Resurrection, other dates all rep
resenting mere inferences from this : " Here was to every Christian eye from
the first the turning point of the world's evolution" ; and the Church's Con
fession had always put in the forefront the historical setting « under Pontius
Pilate" (1 Tim. 6 ia). Turner might well have added a reference to the
annual commemoration of the Redemption by Quartodecimans since apostolic
times.

132 JOURNAL OF BIBLICAL LITERATURE
under 7 on the agreement of Luke's sources with Mt. 2 16) ;
it conflicts seriously, however, with Lk. 2 2, where the census
at the time of Jesus' birth is declared to be the famous one
again referred to in Ac. 5 37, " the first census, the one made
when Quirinius was governor of Syria " (Greek). All expe
dients for identifying this census with some statistical meas
ure " in the days of Herod the king " are futile, since the
census in question was for the Jewish world the epoch-mak
ing event of the century, when Judaea lost the last semblance
of independence. This measure of incorporation into the
empire was carried out according to Josephus by Quirinius,2
after the deposition of Archelaus in a.d. 6-7. It was re
membered not only for the desperate revolt it provoked
(referred to in Ac. 5 37), but for the bone of perpetual con
tention it furnished thenceforth to Jewish casuistry. " Is it
lawful to pay the census to Csesar ? " 3 To suggest a possible
employment of Quirinius in Syria in some other capacity
than "governor," since his predecessors, M. Titius, C. Sen-
tius Saturninus, and P. Quintilius Varus, cover by their ten
ures the whole period from before B.C. 9 till after the death
of Herod, and thus to make room for another unknown census
taken while Herod was king, is a mere subterfuge. On the
other band, Zahn and Spitta have recently shown * that Jose
phus rather than Luke may be the author who is here prin
cipally in error. Revolts took place in B.C. 4-2 in all parts
of Herod's dominions, that in Galilee being led by a certain
Judas, whom Schurer declares to be "certainly identical"
with the famous leader of the revolt "in the days of the
Census." Quirinius is known to have been proconsul of
Syria in B.C. 4-2, but not at the time alleged by Josephus.
In fact, Josephus is particularly ill-informed as to the period
of the census. Whether, then, Luke be right or wrong in
2 Ant. XVII. xiii. 5 and XVIII. i. 1. See, however, the suggestion of
Spitta below, as to a possible error of Josephus regarding the date of Qui
rinius' appointment.
8 Mk. 12 14.
4 Cf. Zahn, " Die Syrische Statthalterschaft u. d. Sehatzung d. Quirinius "
in JV. Kirchl. Zts. 1893, and Spitta, " Die chronol. Notizen," etc., in Zts. f.
ntl. W. 1906.

BACON : NOTES ON GOSPEL CHRONOLOGY 133
connecting the name of Quirinius with the census, and
whether it be allowable or not to think of the census as oc
curring in B.C. 3-2, the mention of the proconsulship of Qui
rinius indicates B.C. 3-2 as the date really intended.8
If this explanation be found inadmissible, there is no in
superable difficulty in attributing to an author guilty of the
glaring anachronisms of Ac. 5 36 f. the antedating by some ten
years of the census of Quirinius. It certainly did not occur
"in the days of Herod the king." The explanatory way in
which Lk. 2 2 is interjected suggests that the whole effort to
connect the story of the birth in Bethlehem with Quirinius
and the famous census is only a somewhat blundering attempt
to connect with outside history, similar to that of Ac. 11 28,
where the local Palestinian famine of A.D. 46^47 is univer
salized. 3. Duration of the Ministry. The synchronisms of Lk. 3 1,
so elaborately constructed, yet followed by no date for the
Passion and Resurrection, nor any intimation that more than
a year expired before this culmination of the drama of redemp
tion, suggest that Luke in common with practically the entire
body of the Christian church down to the time of Eusebius,
understood literally the reference to " the acceptable year of
the Lord " (Is. 61 2) placed in the mouth of Jesus in his
programmatic address (Lk. 4 19). At least there is no indi
cation in the third gospel of departure from the general syn
optic type derived from Mark, in which events lead up to
the final Passover, as if all fell within the limits of the single
ecclesiastical year.
Attempts to determine the duration of the ministry from
indirect indications of the evangelists should be distinguished
from determinations of their own conception. It is possible
and even probable that the evangelists' conceptions had in
course of a generation or more become affected by the primi
tive practice of observing the 14th Nisan as the anniversary
6 Since the above was in type an article has appeared in the Zts. f. ntL
Wiss. x. 4, 1909, from the pen of Rev. Wm. "Weber of Pittsburgh, Pa.,
supporting the view of Spitta, and offering the conjecture that the name
"Sabinus" in Ant. xvii. 221 ff. represents a misreading by Josephus of his
Aramaic source (Nicholas of Damascus), which had Quirinius.

134 JOURNAL OF BIBLICAL LITERATURE
of the Lord's death. This would tend to give to the gospel
drama a cyclic character, condensing into the limits of a
single ecclesiastical year events which really exceeded them.
But the data commonly cited in favor of its longer duration
fall far short of their object. They not only fail to remove
the discrepancy between synoptic and Johannine tradition
(see below, under 8), but even absolutely they do not require
much extension of the synoptic single year. The cyclic con
ception would not be dominant at so early a period if the
ministry had reahy extended over much more than a year,
and 18 months are quite sufficient to account for known
events. Mk. 2 23 (not in chronological order) and 6 39 may
be taken to imply a Passover season spent in Galilee. The
fact that Jesus has friends in Persea and Judaea (Lk. 10 38
Mk. 11 2 14 3 15 43 (?)) has no bearing on the case unless all
Jesus' friendships are to be dated after the opening of the
ministry. Lk. 13 7 is wholly inapplicable, and equally the
quotation Mt. 23 3t-39 (cf. Lk. 11 49), where the personified
" Wisdom of God " complains in a figure derived from
Ps. 91 4 that her repeated appeals to Jerusalem through
" prophets and ' wise men ' and scribes " have been in vain.
All this, accordingly, does not carry us beyond the fact that
Jesus' public ministry extended probably over two Passover
seasons. It does not warrant us in assuming that Luke or
even Mark so understood.
4. Patristic Dating. The Lucan chronology, with which we
are now dealing, starts from some point in the fifteenth year
of Tiberius (April a.d. 28-Aug. a.d. 29), placing the
Nativity 30 years before, and the Passion 12 months after.
This chronology became dominant by the middle of the
second century, especiahy among Gnostic exploiters of the
symbolism of the numbers 30 and 12.6
Even in the particular of the one-year ministry it
remained unaltered for a fud century after the general ac
ceptance of the fourth gospel with its three Passovers
(2 13 ff. 6 4 11 55 ff.) made modification seemingly unavoidable.
Against opponents of the fourth gospel, who pointed to this
6 Irenseus, Her. II. xxii,

BACON: NOTES ON GOSPEL CHRONOLOGY 135
contradiction, Apolhnaris of Hierapolis, and perhaps Melito
of Sardis, had maintained a two-year ministry. Irenaeus
(a.d. 186) attempts to harmonize by the assumption of a
ministry of 20 years' duration. Later harmonizers perhaps
employed thus the traditional 12 years before the Dispersion of
the Twelve (below, ii. B, 3). " Tradition " in Eusebius' time
had settled down to the explanation, which he himself adopts,
that " the three evangelists recorded only the deeds done by
the Savior for one year after the imprisonment of John the
Baptist, . . . but John in his Gospel records the deeds of
Christ which were performed before the Baptist was cast into
prison. . . . One who understands this can no longer think
that the Gospels are at variance with one another." (HE,
III. xxiv.) The great Paschal Controversy, whose first out
break (a.d. 154) coincides with efforts of the learned world
to readjust the systems of dating, led to many attempts at
absolute chronologies, of which the paschal cycle of Hippoly
tus at Rome (a.d. 224) was the ultimate outcome. These
endeavor to carry back an exact reckoning of all gospel events
to the very day and hour of the Nativity. One in particular,
traceable to Alexander of Jerusalem (a.d. 218), and claiming
to go back to " apostolic documents," 7 attempts not merely
to fix by consulships the date of the great ecclesiastical anni
versaries, but also to harmonize the Lucan chronology with
the supposed requirements of John.8 Omitting the days of
the month the datings are as follows : Nativity, A.D. 9 ;
Baptism, a.d. 46; Crucifixion, a.d. 58(!); Resurrection,
Mar. 25,9 a.d. 58.
In all later attempts, and nearly all of the second century,
7 An early form of the Acts of Pilate is probably meant. The Cappado-
cians, from among whom Alexander came, kept the (Christian) Passover on
the vernal equinox of the Julian calendar (Mar. 25 ; others Mar. 18, the date
of the sun's entrance into Aries). This fixation of the anniversary of the
feast of the Redemption on Mar. 25 is the main point of the Alexander frag
ment. But the Cappadocians adduced as their authority for the practice
the Acts of Pilate.
8 Dobschutz, T. u. U. xi. 1, pp. 136 ff.
» On the significance of this date see below, B 2. Mar. 25 fell on Satur
day in a.d. 58, a discrepancy in the source.

136 JOURNAL OF BIBLICAL LITERATURE
the fixed point is the so-called " year of the two Gemini "
for the year of the Passion. It is from this, reckoned as
Tib. xvi, that in Turner's opinion the synchronisms of
Lk. 3 l are calculated. Manifestly the practice of observ
ing the anniversary of Jesus' death and resurrection as a
Christian feast of redemption coincidently with the Jewish
Passover, a practice amply attested as in vogue since the
times of the apostles, if not alluded to by Paul himself
(I Cor. 5 7 15 20), would make the determination of the
year from which this so-called " Quartodeciman " practice
originated a matter for precise determination at a very early
period. 5. Relation to John. More important support for the
Lucan chronology than the adhesion of the second century
fathers may be found in Jn. 2 20. In spite of some ancient
authorities 10 who apply the reference to Solomon's or Zerub-
babel's temple, it seems more probable that the author has
carefully reckoned out the synchronism of 46 years com
pleted since the beginning of Herod's enterprise (B.C. 20-19,
Turner), making the ministry begin with the Passover of 26
and end with that of 29. If with Irenseus we take the feast
of Jn. 5 l as a Passover, as well as 6 4, the duration of the
ministry will be reckoned at three years and Jesus' age as 49.
This author's divergence in other respects from synoptic tra
dition (below, D) cannot therefore affect his clear acceptance
of the main factor in the Lucan chronology, the year 29 for
the Crucifixion. Whether this acceptance is due to inde
pendent tradition, or merely ranks the writer of Jn. 2 20 with
later chronographers who depend on Luke, must be decided
by our judgment of the value of the date, and of the two
systems respectively. It stands, at all events, as a confirma
tion of 29 as the traditional year of the Crucifixion and cor
roborates our interpretation of Lk. 3 l.
We must inquire later as to the duration of the ministry,
and of Jesus' life, implied in Jn. 2 20, and in the fourth
gospel as a whole. 10 Gosp. of Nicod. and De Mont. Sina et Sion.

BACON: NOTES ON GOSPEL CHRONOLOGY 137
B. Criticism op the Lucan Chronology
In spite of this array of second century support, very
serious reason exists to question the accuracy of the date
Tib. xvi (i.e. the year of the two Gemini) implied in
Lk. 3 1, independent of the evidences of this author's super
ficiality elsewhere. The mere fact that ancient and modern
chronographers have not ventured further from it than to
extend the duration of the ministry in the interests of
harmony, or to make some other reckoning of the years of
Tiberius is not a guarantee ; and if the pivotal date of Lk. 3 1
with the connected " year of the two Gemini " be invalidated,
the whole chain of derived chronologies falls with it.
1. The Calendar Argument. It is a statement of all the
Gospels, of practically universal acceptance,11 that the Cruci
fixion occurred on a Friday. The Gospels differ, however,
as to its having occurred before (John) or after (synoptics)
the Passover supper, i.e. on the 14th or 15th of the luna
tion Nisan, which was invariably that of the first full moon
' after vernal ^eqi^ngx. Astronomical calculation can deter
mine within limits of reasonable certainty on what years of
Pilate's administration Nisan 14 fell on Thursday, which
would make the year possible from the point of view of the
synoptic writers; or on Friday, which would agree with
John. No year in which Nisan 14 fell neither on Thursday
or Friday can possibly have been that of the Crucifixion.
Naturally this method of determination by exclusion has
been repeatedly applied, but with varying results, not from
uncertainty in the astronomical calculations, which determine
with exactness the day and hour of what should have been
the Passover full moon, but from uncertainty as to the degree
of approximation to this date with which the feast was ap
pointed by the Jewish authorities. It admits, however, of
positive demonstration ffi that the day of the month was fixed
11 An exception must be made of E. Schwartz (« Osterbetrachtungen,"
Z. nt. W. vii. 1, 1906, pp. 37 ff.), who considers that the observance of " the
Lord's day " as the resurrection day and of Wednesday and Friday as fast
days (AiS. viii. 1) arose independently of the events of Passion week.
12 Preaching of Peter ap. Clem. Al., Strom, vi. 6 ; cf. Bacon, "Incidence
of Passover," in Expositor, July, 1900.

138 JOURNAL OF BIBLICAL LITERATURE
by actual observation of the moon's phasis at the beginning
of the lunation. The day preceding that night in which two
credible witnesses had seen the slender crescent of the new
moon was " sanctified " on the following day (second of the
month), according to the requirements of the Talmudic
treatise on " The Beginning of the Year," as the first of the
month. The new moon of Nisan (first month of the year)
was, of course, the most important. Several uncertainties
remain, however, to limit our process of exclusion.
a. Whether a given lunation should be regarded as the
first of the new year (Nisan) or added (by intercalation) to
form the last of the old (Veadar), would admit of doubt
only when full moon nearly coincided with spring equinox.
Jewish astronomy was not equal to the task of determining
the equinox more nearly than within one or two days, and
the rules given for choice of the lunation Nisan are therefore
quite crude and empirical. In such ambiguous years two
possible lunations must be calculated, either of which may
have been Nisan.
b. Cloudy weather might delay the "sanctification" of
Nisan 1. The delay, however, could not exceed 24 hours,
because of the rule limiting the months to 30 days.13 We
must therefore include in our calculations both the 14th and
15th Nisan.
c. The Synoptic Gospels represent the Crucifixion as
taking place on Nisan 15 ; John, with greater probability
(below, D, 2), on Nisan 14. Two days must be admitted
as possible in each Nisan on this score. Fortunately, the
adowance largely cancels that of b ; for a delayed Nisan 14
would coincide with Nisan 15.
With these allowances to cover all uncertainties it becomes
possible to rule out certain years, in which, on no admissible
supposition, either Nisan 14 or 15 could have fallen on a
Friday. 18 Turner, ibid. p. 412, declares it as " one of the rules which may well go
back to our Lord's time that Adar never consisted of more than 29 days."
This would still further reduce the elements of uncertainty our tables are
designed to cover, but the authority is not stated.

bacon: notes on gospel chronology

139

The subjoined table, taken substantially from Fothering-
ham (Journ. of Philol. 1903, p. 100 ff.), avoids the crudity
of those previously constructed on the uniform average
allowance of 36 hours between astronomical new moon and
phasis.14 By determining the astronomical conditions of
visibility in each lunation in terms of the moon's Mean
Anomaly and Latitude of the Ascending Node, it fixes with
practical certainty the evening before which the moon was
invisible. By including the years 26 and 27 and calculating
a second date for 26, 29 and 34, in which Nisan 14 would
have fallen before Mar. 25, the equinox of the Julian cal
endar,16 two further sources of uncertainty are removed.

Adab

(Veadak) Nisan

Astr.

New

'

Astr.

New

A.D

Moon

Phasis

Moon

Phasis

Nisan 14

26

6 Feb.

8 A.M.

8 Feb.

f 7 Mar.,
{ 6 Apr.,

7 P.M.
6 A.M.

9 Mar.
7 Apr.

Sa. 23 Mar. 1
Su. 20 Apr. j

27

25 Feb.

5 A.M.

27 Feb.

26 Mar.,

6 p.m.

28 Mar.

Fr. 11 Apr.

28

14 Feb.

9 A.M.

16 Feb.

15 Mar.,

2 a.m.

17 Mar.

We. 31 Mar.

29

2 Feb.

8 A.M.

4 Feb.

f 4 Mar.,
1 2 Apr.,

3 A.M.
8 P.M.

6 Mar.
5 Apr.

Su. 20 Mar. '
Tu. 19 Apr.

30

21 Feb.

5 A.M.

23 Feb.

22 Mar.,

8 P.M.

f 24 Mar.
[ 25 Mar.

Fr. 7 Apr.
Sa. 8 Apr.

31

11 Feb.

0 A.M.

12 Feb.

12 Mar.,

1 A.M.

14 Mar.

"We. 28 Mar.

32

29 Feb.

1 P.M.

2 Mar.

29 Mar.,

11 P.M.

31 Mar.

Mo. 14 Apr.

33

18 Feb.

5 A.M.

19 Feb.

19 Mar.,

1 P.M.

f 20 Mar.
{ 21 Mar.

Fr. 3 Apr.
Sa. 4 Apr.

34

7 Feb.,

8 P.M.

9 Feb.

f 9 Mar.,
1 7 Apr.,

6 A.M.
2 p.m.

10 Mar.
8 Apr.

We. 24 Mar.
Th. 22 Apr. /

35

26 Feb.,

8 P.M.

28 Feb.

28 Mar.,

6 A.M.

29 Mar.

Tu. 12 Apr.

The only admissible years among the above, if synoptic
tradition be right, are 34, in which, if the later lunation was
proclaimed Nisan, the 15th fell normally on Friday ; and 28
and 31, if bad weather delayed the phasis. But no amount
of delay from this cause would produce any effect at the
beginning of Nisan in 28 and 31, because Adar, the preced
ing month, had its full quota of 30 days. Only on the sup-
14 So Wurm, Anger, Wieseler, and Achelis.
" Wurm's rule (ap. Wieseler, tr. Venables, Syn. of Four Gospels, p. 406),
giving Mar. 23 for the equinox, is too exact. The Sanhedrin may have fol
lowed ordinary (Julian) usage.

140

JOURNAL OF BIBLICAL LITERATURE

position of delay both at the beginning of Adar and of Nisan
do these years become admissible. They are not admissible
at all unless synoptic tradition be right as against Johannine.
If Johannine tradition be right, 34 is again the most
probable year; for while the conditions of visibility were
such as to make Thursday, Apr. 22, a.d. 34, a possible
14 Nisan, thus confirming synoptic story, this is only just
possible, and a single day's delay in observation from clouds
or otherwise would make it Friday, Apr. 23, confirming the
fourth gospel. Two days' delay (in this case quite sup-
posable) would admit the date of the alternative Nisan.
For in this year, a.d. 34, Friday, Mar. 26, would then have
been Nisan 14, again confirming the fourth gospel.
But besides the year 34 (the most probable by either Johan
nine or synoptic tradition), the years 33 and 30 are both
admissible, though by Johannine tradition only. Fother-
ingham adopts 33 as the true date, because " the date Satur
day, Apr. 4, for Nisan 14 in this year is astronomically only
slightly more probable " than Friday, Apr. 3. Of 30 he says:
" In the case of 30 the conditions are so pronouncedly in
favor of a late phasis that it would be difficult to adopt an
earlier date " than Saturday, Apr. 8.16 This would of course
exclude 30 a.d. from consideration. Nevertheless, Friday,
Apr. 7, must be admitted to be possible, as cases are known
of observation of the new moon under conditions more un
favorable. We therefore include this as a possible date, and
30 as an admissible year under Johannine tradition. Under
synoptic it is entirely inadmissible.
As regards 29, the patristic " year of the two Gemini " cor
responding to Lk. 3 1, Fotheringham's discussion is exhaustive
and his conclusion decisive. " Whatever date therefore be
astronomically possible, 29 appears to be impossible." Even
Achehs, who reverts to the demonstrably false method of
18 In the table as given above, the columns giving Mean Anomaly and
Longitude of the Ascending Node have been omitted. 200 marks the maxi
mum of unfavorable conditions which become more favorable as the number
advances toward 400 or recedes toward 0. At Nisan a.d. 30 they were 164
and 71 respectively.

BACON: NOTES ON GOSPEL CHRONOLOGY 141
dating by astronomic new and full moon,17 makes Nisan 14
in a.d. 29 fall on Sunday, Apr. 17, whether by this method
or in calculating by phasis!18
2. Source of the Lucan Date. If then the patristic " year
of the two Gemini " be really inadmissible, how account for
the origin of the dating ? In the view of Turner it does not
rest upon Lk. 3 l, but antedates the third gospel itself. At
all events, Luke has some source for it, because it cannot be
connected with any other datum of his story. Whosesoever
the responsibility, the explanation of the adoption of this date
is not far to seek. It is suggested by the course actuahy
pursued by Hippolytus in dating the Crucifixion on Mar. 25,
A.D. 29.
Epiphanius (Hcer. 1. 1, L. 1, 23) tells us of primitive Quar-
todeciman sects in Cappadocia who, in continuing their ob
servance of Passover as the anniversary of the Crucifixion,
had broken away from the Jewish lunar calendar and, instead
of observing Nisan 14, observed the solar date, Mar. 25, just
as we celebrate the Nativity at winter solstice, Dec. 25,
observing the same day of the (solar) month each year.
The certainty that the date, Mar. 25, was an arbitrary one
is afforded by the practice of some of their number, who
observed not Mar. 25, but Mar. 18 ; for in both cases we have
simply spring equinox of the Julian Calendar, where Mar. 18
is the date for the entrance of the sun into Aries, Mar. 25 true
equinox.19 Now in a.d. 29, Mar. 18 and Mar. 25 actually
fell on Fridays. In all other conceivable years equinox fell
on some other day of the week ; so that it is quite suppos-
able that A.D. 29 was made the basis of the synchronisms
17 Turner and Ramsay are certainly wrong in this assumption. Just because
the Jews considered accuracy in the determination of their festal calendar
vitally important, they clung with fanatical tenacity to their determinations
by actual observation long after Gentile astronomy was competent to fix the
dates by calculation. See the fragment of Ker. Petri, quoted above, note 12.
18 G'ott. gel. Nachr., phil. hist. Kl., 1902, pp. 707 ff. The obsolete method
of allowing a uniform delay of 36 hours between astronomic new moon and
phasis accounts for the discrepancy with Fotheringham's results.
19 See above on the fragment from Victorinus discussed by Dobschutz
(T. u. TJ. xi. pp. 136 ff.), in which this dating for the Quartodeciman anni
versary is attributed to Alexander, the Cappadocian bishop of Jerusalem,
ea. 210 a.d.

142 JOURNAL OF BIBLICAL LITERATURE
of Lk. 3 1 for this reason, the dating of Jn. 2 21 and of the
patristic chronographers simply resting upon the same. For
Quartodecimans must very soon have discovered the remark
able coincidence that the Friday of Passover week in " the
year of the two Gemini " coincided with the (Julian) vernal
equinox, just as Jesus' birth was later understood to coincide
with the (Julian) winter solstice. At all events, 29 a.d. as
the year of the Crucifixion seems to be the one universal,
unanimously attested date of all the earliest sources. Every
dating takes its start from this — and yet it is astronomi-
cady impossible!
Per contra, since the year 30 is not altogether excluded
under Johannine tradition, and second century interpreta
tion of Lk. 3 1 may have been mistaken, the possibility
must also be recognized that Lk. 3 1 may be based on inde
pendent historical tradition. Jn. 2 21 will then no doubt
be adduced as supporting the same. In this case, however,
the two passages must be understood as pointing to a.d. 30,
and not 29. A decision as between the alternative possi
bilities must await further evidence.
C. Pre-Lucan Chronology
1. Traces in Matthew and Luke. On other points than the
universally accepted starting point of 29 A.D. for the Cruci
fixion there is decided disagreement in the sources. They
vary both as to the duration of the ministry and of Jesus'
hfe. We have seen above (A, 2) that Luke's calculations of
date do not always agree even with his own material. In
particular the Nativity is placed " in the days of Herod the
king," by Lk. 1 5, if indeed the original reference of Lk. 2 1
be not to some purely statistical census made under Saturni-
nus in B.C. 9-7. The implications of Mt. 2 16 (note last
clause) make 6 B.C. the lowest limit, but a date several years
earlier may quite as well be implied.20 Thus the 30 years of
20 If the legend of the star Mt. 2 2 . be really an outgrowth of the remark
able conjunction of Jupiter and Saturn in Pisces in May, October, and De
cember of b.c. 7 in accordance with the statement of R. Abarbanel that the
conjunction of these two planets in Pisces is to be a sign of Messiah's coming,
we may be still more specific. A date in the neighborhood of b.o. 7 will have
been assumed by the tradition.

BACON : NOTES ON GOSPEL CHRONOLOGY 143
Lk. 3 23 give in any case a lower estimate of Jesus* age " when
he began " than the older tradition implied. We have evidence
which is far from deserving the neglect it has received that
his age was in reality more nearly 40.
2. The Elders' Tradition of Jesus' Age. Irenaeus, in his
attempts to defend the fourth gospel against the charge of
unauthenticity,21 refers to an author, now generally acknowl
edged to be Papias (ca. a.d. 145), as reporting from " the
Elders who were disciples of the Apostles," that Jesus when
he began to teach had reached the age of 40 years, the age
which "befits the teacher." We may infer from this latter
expression that in its original form the tradition was a vin
dication of Jesus' right to the title and function of Rabbi
under the legal requirement of 40 as the teaching age.22
Irenseus twists this into a defense of his interpretation of
Jn. 8 67, as implying that Jesus continued his teaching until
nearly the age of 50 ! Regarding this latter conception (that
Jesus attained the age of 50 years) we shall have more to say
presently. That of Papias with which we are now dealing
(40 years) is not only reasonable in itself, and quite in har
mony with Mt. 2 16 Lk. 1 5, but has positive confirmation in
Ac. 7 23, where the statement is made that Moses when he
visited his brethren in the role of peace-maker, " supposing
that they would understand how that God by his hand was
giving them salvation," had "fully completed his fortieth
year."23 This trait in Stephen's portrait of Moses the great
prototype of Christ is introduced without more than inferen
tial warrant in the Old Testament, which would hardly be the
case if it were not intended to mark a correspondence.
3. Jewish Polemic. Conversely, Herford adduces evidence
from the Jewish side of this polemic between Church and
Synagogue which indicates Jewish interest to prove that
Jesus did not attain the age of 40. Jesus is referred to (as
21 Her. II. xxii. 4, 5.
22 Aboda Zara, b. Talm. Frankfort ed. f . 19 6. Ad quodnam vero aetatis
momentum expectandum est antequam vir doctus alios docere possit ?
Mesp. Ad exactos annos quadraginta.
23 ^ttXtjppCto airQ TctrcrepaKOrraerfy xpbvos.

144 JOURNAL OF BIBLICAL LITERATURE
Herford shows) in Bab. Talm., Sanh. 106 6, under the mask
of " Balaam " (slain for seducing Israel, according to Num.
31 8. 16). A Christian opponent of Rabbi Hanina answers
the question " How old was Balaam when he died ? " as " 33
or 34 years." This is justified thereupon by combination of
Ps. 55 23 and Ps. 90 io. Because Jesus was a " bloody and
deceitful man " he must have perished before attaining half
a man's allotted days (i.e. 70 years). The (Palestinian)
elders on the one side, rabbis on the other, corroborate the
existence of a tradition independent of the Lucan, fixing the
age of Jesus not at thirty, but at forty.
B. Johannine Chronology
We are thus brought again to the question so drastically
treated by Irenaeus. What is the relation of the Johannine
chronological system to the Lucan ?
1. The Johannine Synchronism. Jn. 2 13-22 transfers the
scene of the temple-cleansing from the end of the ministry
(Mk. 11 15-18 and paraUels), its only possible historical posi
tion, to the beginning, thus making it occupy a position
corresponding to that of Lk. 4 i6ff.; where Luke depicts the
beginning of the ministry by describing how Jesus proclaimed
" the acceptable year of the Lord " in his irarpk. This trans
fer with the connected indirect synchronism (vv. 20. 21) and
the emphasizing in 6 4 of a point only indirectly recogniz
able in Mk. 6 39, suggests that the fourth evangelist, while
accepting the Lucan fixed point of A.D. 29 for the Crucifixion,
has a deliberate, if tacit, intention to readjust the prevailing
chronology (a) in respect to the age of Jesus, (J) in respect
to the duration of the ministry, (c) in respect to the day of
the Crucifixion.
a. Early errorists in Rome referred to by Augustine, M per
haps Caius and the Alogi, insisted on interpreting not only
Jn, 8 57, but also 2 20, as implying that Jesus attained an age
of nearly 50 years. Irenaeus, as we have seen, the most deter
mined opponent of the Alogi, admitted this interpretation of
Jn. 8 57, and other Church writers, not long after, admitted it
24 de Doctr. Chr. ii. 28.

BACON: NOTES ON GOSPEL CHRONOLOGY 145
as regards 2 20.25 We stdl have, in fact, many traces of bar-
monistic chronologies of this period which reckon the dura
tion of Jesus' life at 49 years.26 As these interpretations
did not disappear until their real irreconcilability with Luke
became apparent, so it is probable that but for prepossessions
of a similar kind we ourselves would recognize that at least
Jn. 8 57 is not written from the standpoint of Lk. 3 23, but
more nearly from that of the pre-Lucan chronology of Mt.
2 ie, Lk. 1 5 2i (?) Ac. 7 23 and "the Elders" of Papias, if,
indeed, John does not go still further in the direction of
Irenseus' extravagant claims.
b. We have already recognized the probability that Jn.
2 13-22 is intended to rectify the Lucan thirty-year duration
of Jesus' life ; we must now recognize the further fact that
this writer certainly extends the duration of the ministry,
and the probability that he intends it to cover a period of
three years, perhaps taking the suggestion from Lk. 13 7.
We may be sure from Jn. 8 57 that he reckoned the dura
tion of Jesus' life as at that time (Tabernacles) approximating
50 years. But did he think of it as attaining just 50 at
the ensuing Passover ? If he reckoned after the manner of
the later Greek chronographers, his starting point for the
period of the incarnation would be not the birth of Jesus, but
his conception (o-wWij^rts) in the womb of Mary. " Forty-
six years" of age, therefore, would mark more nearly 47
26 So among moderns Loisy and Julicher.
26 A number of medieval chronographers who seem to rest upon the Defense
ofthe Fourth Gospel and Apocalypse, by Hippolytus (ca. a.d. 205), calculate
by jubilees of years, reckoning 120 jubilees (6000 years) as the duration of
the world, after which the millennium completes the period of 140 (2 x 70).
Thus a fragment of the Apocalypse of Thomas quoted by Frick (Zts. f. ntl. Wiss.
1908, 2, p. 172) forecasts "nine jubilees" (=450 years) from the Ascension
to the Second Coming. The Nativity is dated in chronologies of this class in
a.m. 5500 (so, e.g., Annianus, ca. 412 a.d.), so that Apoc. Thorn, probably
reckoned one jubilee (50 years) as the period of the Incarnation. Chapman,
however, has cogently argued (Journ. of Theol. Studies, viii. 1906) for the
dependence of this reckoning on Hippolytus, who employed the blundering
calculations of Tertullian from the consular lists. Thus the extraordinary
dates of Epiphanius and the Alexander fragment giving 12 years as the
duration of the ministry and 49 as the duration of Jesus' life would be purely
fortuitous in origin 1

146 JOURNAL OF BIBLICAL LITERATURE
of the incarnation period. We may also assume that he
regarded the unnamed " feast of the Jews " of Jh. 5 1 as a
Passover, because such is the general understanding of the
fathers.27 If, then, a correspondence is intended between
Herod's building and " the temple of his (Jesus') body," the
Passover of the utterance will approximate the close of the
forty-seventh year of the incarnation, that of 5 1 the forty-
eighth, that of 6 4 the forty-ninth, and that of the Passion and
Resurrection the fiftieth year.
On the other hand, the internal structure of the fourth
gospel convinces us that Turner is right in his declaration
that it contemplates originally and fundamentally no other
chronology of the ministry than is reasonably and most
naturaUy to be derived from Mark, viz. a period of one year
for the Gahlean ministry, ending with the Passover of Ch. 6,
and a second, also of one year, for the ministry in Persea and
Judsea, ending with the Passover of Ch. II.28 In this case
Jn. 2 13 «. wih represent an extension of the original Johannine
duration of Jesus' ministry, as wed as an attempt at more
exact dating of his birth.
c. The placing of the Crucifixion on Nisan 14 (Jn. 13 l. 29
18 28), instead of 15 as the synoptists declare, is an extraor
dinary instance of the combined boldness in matter with
reserve in manner of the corrections of our fourth evangelist.
His undeniable disposition to find mystical coincidences (e.g.
9 7 11 50) has indeed been regarded as sufficient to account for
even such a tour de force as this, inasmuch as Jesus' death,
the author's parallel to that of the Passover lamb (1 29. 36), is
thus made to fad coincidently with this sacrifice. But it
is to be observed that the real author of this comparison is
Paul (1 Cor. 5 7), and that Paul adds to it a further com-
27 In our own judgment a misunderstanding. The dialogue of Jn. 5 is
concerned with the authority of Jesus vs. the Law of Moses. Pentecost was
the feast of the giving of the Law. But this only shows that Jn. 2 13 a. is a
late addition, which can be made probable independently.
28 Turner, s.v. "Chronology of N. T." in Hastings' BD, i. p. 409 6).
" St. Mark's Gospel was shown (p. 406), if its order of events can be taken
as chronological [the fourth evangelist would so take them], to imply,
exactly like St. John's, a two-year ministry."

BACON: NOTES ON GOSPEL CHRONOLOGY 147
parison of the Resurrection to the waving of the Firstfruits
(1 Cor. 15 20), which took place on Nisan 16. This he
would scarcely do if the days of Crucifixion and Resurrection
had fallen respectively on Nisan 15 and 17. Moreover, there
is strong internal evidence of self-contradiction in the story
of Mk. 14 ff . and parallels, indicating that the temptation to
regard the last supper as the actual Passover with which it
so nearly coincided, has availed to alter the original repre
sentation. Even now the account is bare of all the distinc
tive features of the Passover meal. In all probability, then,
we must invert the supposed relation. Symbolism is unques
tionably an exaggerated factor in John; but it grows out of
the historic tradition, and not conversely. And what we
thus find to be true respecting the day of the Crucifixion is
at least partially true of the duration of the ministry, which,
if not exactly two years, was almost certainly more than one;
also of Jesus' age, which if not exactly 49, was very probably
in the region of 40.
E. Harmonistic Chronologies
Current harmonistic chronologies based on a 3-4 year
ministry date from the period of Eusebius, who adopted 29
(Lk. 3 l) as the date of the Baptist's mission, 30 as that of the
Baptism, and 33 for the Crucifixion, finding four Passovers
in John. The heathen chronographer Phlegon, A.D. 117-138,
furnished him with records of an earthquake which had
occurred in 32-33 in Bithynia, and an eclipse of the sun
(certainly that of Nov. 24, A.D. 29) visible in the same
region. Unfortunately there is no good reason to suppose
tbat any but Christian borrowers from Phlegon identified
these events with the " eclipse " of Lk. 23 44 and the earth
quake of Mt. 27 51. Still it is of interest to note Phlegon's
date for the earthquake, if not for both events, in view of the
fact that other Christian chronologists before Eusebius may
have brought it into relation with gospel story. Phlegon's
date was 01. 202. 4 = a.d. 32-33.

148 journal of biblical literature
Conclusion
The date a.d. 29 for the Passion seems to be older than
any of our gospels, and yet on the showing of astronomy
both incorrect and factitious. Luke would seem to have
employed it for his dating of the beginning of the min
istry "in the fifteenth year of Tiberius." The writer of
Jn. 2 13 ff. would seem also to have calculated from it ; but
whereas Luke — misled perhaps by the desire to connect the
birth in Bethlehem with " the census " — placed the Nativity
but 30 years before it, John, more loyal to the older tradi
tion of Jesus' having "attained the age of the teacher,"
placed the Nativity more than 40, perhaps exactly 49 years
before it. Astronomically the most probable dates for the
Passion are A.D. 33 or 34. The most probable estimates of
the duration of the ministry are those which come nearest
the two-year period. The most probable estimates of the
duration of Jesus' hfe are those which exceed rather than
fah below the period of 40 years. The Nativity should be
dated " in the days of Herod the king," most reasonably in
about B.C. 8.

6964