Alexandrian judaism

Josephus only once mentions Philo, in a brief notice of the embassy to Caligula, which was led by the philosopher to oppose the counter-embassy of Apion (Ant. XVIII. 257 ff.) He there speaks of him in the highest terms a8 dvip ra wdvra evdotos .. kal didocodias ov Gmrecpos.

It is impossible to say whether Josephus was acquainted with the detailed account of that embassy which Philo has left us in his Legatio ad Gaium, or how far he was acquainted with the other writings of the Alexandrian philo- sopher. Had he accomplished his projected work on the Being of God and the Meaning of the Laws, we should be in a better position to estimate the extent of the influence which Philo exercised upon him.

Indications, however, are not wanting in the early books of the Antiquities of an appar- ently direct dependence upon Philo’s writings. The following are the principal parallels which have been noted :—(1) The Preface to the Anti- quities and the opening of the de Opijicio Mundi show a striking agreement in the sequence of ideas. Both works raise the question why the Mosaic code is preceded by an account of the Creation.

Josephus expects that his readers will wonder how it comes to pass that his work, of which the main purpose is to record laws and historical events, has so large an element of ‘physiology’ (ért rocovroy gduciodoylas Kexowdvnker).

He explains that Moses, differing in this respevt from other legislators, whose codes begin with contracts and the rights of man, considered it necessary, before laying down his code, first te elevate men’s minds by setting the highest of all examples before them and inducing them to con- template the nature and actions of God, especially as exhibited in the creation of the world (Ané. 1 18 ff.)

Philo begins his work with a similar con- trast between the procedure of Moses and that of other legislators. Moses did not commence by laying down commands and prohibitions, but gave as his exordium a most marvellous account of the Creation, in order to show the harmony existing between the world and the Law, and that the law- abiding man is a true citizen of the world. The unanimity of the Law and the universe is also expressed by Josephus (Ant. I. 24, wdvra yap rp Tov doy pice.

ciupwvov exer Thy didbecw), Josephus (I. 15, 22) and Philo both refer to the mythical stories which disfigure the codes of other legis- lators. (2) In the same context, Josephus, quite in accordance with Philo’s doctrine, admits that there is an allegorical meaning in Scripture as well as a literal (I. 24, ra wey alvirropévou rob vopobérov des, Ta 5 GAAYopoUvTOs peTa gewvdryros, baa S evdeias AéyerOar cuvédepe, Tatra pyres éupavlfovros).

It is not often that Josephus in the Antiquities resorts to such allegorical explanation [that was reserved for the projected airioAoyla]; but there is one striking instance, where the tabernacle and its furniture and the various articles in the dress of the high priest are explained as symbolical of the universe and its parts (Ant. I. 179-187).

This is quite in the style of Philo, who gives a similar interpretation of the materials used for the woven hangings for the tabernacle and the high priest’s apparel, in the de Vita Mosis, iii. 6, 12. The details of the explanation are not abso- lutely identical in the two writers, but for the general idea Josephus is not improbably directly dependent upon Philo. (3) In Ané. I. 29 an ex- planation of the use of ula for rpéry in Gn 1* is promised in the alriodoyla.

For Philo’s explana- tion, see de Opific. Mundi, 9. (4) Some of the explanations of Hebrew proper names are iden- tical in the two writers: these, however, may go back to an earlier tradition. (5) Some expressions with regard to the nature of God have the ring of Philo, or at least of Alexandria. Seec. Ap. ii. 167 (dyévnjrov kal mpds tov aldvov xpbvov dvaddolwrov .. Owdper pev huiv yvwpimor, drotos 6é¢ kar’ ovolav égrlvy dyvworov); Ant. VI.

230 (rdv Gedy rodroy dv modvdv opas Kal mavraxov Kexupuévov), X. 142, 278; c. Ap. li. 284 (6 Beds dud wavrds Tod Kdcpou redolr- nxev). The four cardinal virtues of Greek philo- sophy are traced by Josephus, as by Philo, in the Mosuio code (c. Ap. ii. 170; cf. Wis 8’, with Deane’s note). But the indications which Gfrorer (Philo, 1831, ii.

356-367) has found in Josephus of the Logos doctrine of Philo—in the account of the three angels who visited Abraham, the burning bush, and the pillar of fire—are fanciful and far from convincing. See Siegfried, Philo von Alex- andria, 1875, pp. 278-281. v. THE ALLEGED WITNESS OF JOSEPHUS TO CuRisT. —The passage on which so much _ has been written occurs in Ant. XvIU. 63f. [iii. 3], and runs as follows: ‘ Now about this time lived Jesus, a wise man, if indeed one should call him a man.

For he was a doer of marvellous works, a teacher of such men as receive the truth with pleasure ; and many of the Jews ani many also of the Greeks did he win over to himself: this was the Christ.

And when, on the indictment of the principal men among us, Pilate had sentenced hins JOSEPHUS JOSEPHUS to the cross, those who loved him at the first ceased not [to do so]; for he appeared to them on the third day again alive, as the Divine prophets had declared these and ten thousand other wonder- ful things concerning him. And even now the race (rd @dXov) of Christians, which takes its name froin him, is not extinct.’ The passage stood in the text of Josephus in the 4th cent.

, as Eusebius ances it (HE i. 11; Dem. Ev. iii. 3. 105f., ed. xaisford), and from that time down to the 16th cent. its genuineness was undoubted. Its exist- ence contributed largely to the high esteem in which Josephus was held by the Fathers. During the last 300 years a vast amount of literature has been written on the question of its authenticity.

Very few critics at the present day accept the pas- sage as it stands as from the pen of Josephus; but there is a division of opinion as to whether the whole is an interpolation, or whether Josephus did make a brief statement about Jesus Christ, es was afterwards augmented by a Christian and. (1) As to the external evidence, it is true that the passage occurs in all the MSS. But this is of comparatively little weight, as none of the Greek MSS containing Book xvut.

of the Antiquities is older than the llth century. The old Latin version carries us much further back, to the time of Cassiodorus (beginning of the 6th cent.), and the quotation in Eusebius attests the existence of the passage still earlier, in the 4th century. On the other hand, it is practically certain that Origen in the preceding century did not find it in his text of Josephus.

For, while he is aware of the passage in Josephus concerning James, the Lord’s brother, ‘he says: ‘The wonder is that though he did not admit our Jesus to be Christ, he none the less gave his witness to so much righteousness in James’ (Comm. in Matt. x. 17); elsewhere Origen, collecting all the indirect evidence for Christianity which he can find in Josephus, is silent on the above passage, and again states that Josephus ‘ disbelieved in Jesus as Christ’ (c. Celsum, i. 47).

This is a case where the negative evidence practi- cally amounts to a positive proof that the passage ~Was unknown. (2) The internal evidence is decisive against the genuineness of the passage as it stands. The style affords no certain clue: it is not markedly different from that of Josephus in this part of his work: it may be granted that the interpolator has done his work with someskill. But the contents are not such as Josephus could have written.

He is elsewhere, as was seen, silent on the subject of a Messiah. The sentence ‘this was the Christ’ (#v, not évoultero) ean have come only from a Christian pen, and it is certain that Josephus was not a Christian. The same may be said of the phrases ‘if one should call him a man,’ ‘the truth,’ and the statement about the appearance on the third day. Zahn has adduced an interesting parallel to the first of these phrases and the following words ‘for he was a doer,’ ete.

, from a Christian work, the Acta Pilati (quoted in Schiirer). The passage is out of place, and breaks the sequence of the narrative. It is interposed between an account of the disturbances in Judiea caused by Pilate’s disregard of Jewish Keruples (55-62), and an account of scandals con- nected with the worshippers of Isis and the banish- ment of Jews from Rome (65-84). The opening of xvul1. 65, ‘And about the same time another calamity disturbed the Jews,’ connects that section .

directly with the section about Pilate. The men- tion of Pilate has of course led to the insertion of the passage at this point. The fact that the passage interrupts the sequence of the narrative is an argument for its spuriousness as a whole. Moreover, as Schiirer has pointed out, a careful analysis of the section, eliminating all that must be of Christian origin, leaves practically nothing behind. The theory of partial interpolation is unsatisfactory.

Two other passages have to be taken into account in the discussion: (a) that concerning the death of ‘John surnamed the Baptist’ (Ané. XvuI. 116- 119), who is described as a good man who bade the Jews practise virtue and be baptized, and who was put to death by Herod because he feared that John’s influence over the people might lead to a rebellion ; (6) that concerning the death of James, ‘the brother of Jesus who was called Christ,’ whom Ananus the high priest caused to be stoned (Ant. xx.

200f.) Origen refers to both these passages. There is no reason why the former should not be accepted as genuine. The style is distinctly that of Josephus [N.B. the form apaprds in xviii. 117].

The historian could refer to the preaching and baptism of John without giving offence to his Roman readers; he could not without personal risk allude to Messianic expectations at a time when the spirit of the Jewish revolt against Rome, the strength of which lay in those expectations, had not been completely quelled. The language of the second passage is not inconsistent with its authenticity. There is a marked difference be- tween the words ‘who was called Christ’ and ‘he was the Christ.

’ But since Origen, in referring to the passage (c. Celswm, i. 47), says that Josephus attributed the outbreak of the war to the putting to death of James (a statement which does not occur in our text), there is good reason to believe that here, too, there has been interpolation. This has taken various forms, one of which is that given by Origen. We conclude, then, that the passage about Christ was introduced into the text by a Christian reader towards the end of the 3rd cent.

, between the time of Origen and that of Eusebius. For the literature and an admirable discussion of the question, see Schiirer, GJV* i. 544-549 (to which the present writer is largely indebted). For the passage about James, see i. 581 ff. of the same work. vi. STYLE.—A few remarks may not be out 9%f place with regard to the style of the historian, upon which there can be no doubt that he spent considerable pains. He tells us as much in Ant. Xx.

263; and, while he justly claims to have acquired a certain skill in the grammar, he con- fesses that long usage of his national language had prevented his mastering the Greek pronuncia- tion (riv mpopopdv). Elsewhere, he tells us that his chief aims are accuracy and beauty of style (76 rijs émayyeNlas Kéddoz), so far as this is attainable b the choice of words and their arrangement, an the use of other ornaments of speech (Ant. XIV. 2).

His fastidiousness in this direction may be illus- trated by the way in which, while using the LXX, he regularly replaces certain words used by the translators by others of a more literary character. Thus he uses éc67s for LXX iudriov (-topuds), éumep- mpdvae for éumuplfev, kardmw for kardmbev, veavioxos for maddpiov, méumew for drooré\\ew, dravyray for auvavrav, bwoorpépew for ériorpépew.

Similarly, he has taken the trouble to re-shape most of the sentences in the Letter of Aristeas, while retain- ing a good deal of the language. His Greek is almost entirely free from Hebraisms; the use of mpocridecda (like Heb. 40°) is the only certain instance which Schmidt discovers (de Flav. Jos. Elocutione, p. 516). He tells us that, in writing the Jewish War, he employed collaborateurs to assist him with the Greek (c. Ap. i.

50, xpnoduevds riot mpos Thy ‘DAdyvida puaviv cuvepyors), and no doubt he had similar assistance in writing the Antiquities. Tt would be interesting to know how far their work extended. Naturally, variations in the style JOSEPHUS NUMBERS, HOURS, AND YEARS 473 ‘and vocabulary occur, partly due to the different ‘sources on which he draws, partly perhaps to the advice of different cvvepyol. The most marked instance of change of style occurs in three of the later books of the Antiquities (XVII.

XVII. XIX.) Among the most striking of the phrases and uses peculiar to or characteristic of these three books, the following may be noted : a large use of the neuter participle (pres. aor. pf. fut.) with article as an abstract noun [e.g. Xvi. 1, TO nH Emckowwracor 3 171, ev édrldt Tod dvacdadodyTos ‘of recovery’: a list is given in Schmidt, op. cit. 361-368 : the use is Thucydidean]; a more frequent use of the optative [Schmidt notes that the con- junctive is absent from Book XVII.]

; the use of «i ‘with inf. in oratio obliqua, of the Attic termina- tion -aro for -vro (Thucydidean), of éréc0s where éoos is used in the earlier books, of édc7icotv (rap dyTwodv, overwas=mdvras), and the phrases éx rov d&éos, under els dvaBodds (cf. Thucydides), and, com- bined, udev els dv. ad’ éx rod déé0s. The departure in these books from the ordinary practice of the ‘writer extends to the orthography.

Whereas else- where Josephus, according ta the MSS, almost invariably writes the Attic r7, in these books cc is the rule, and rr is almost unrepresented ; it begins to recur towards the end of Book xIx., and in Xx. the two spellings occur in almost equal proportions.

It must be added that in these books the imitation of Thucydidean words and phrases is more marked ; the writer has tried to reproduce the difficult style and involved periods of his model, with the result that he has often made his meaning very obscure, and the text has suffered much corruption. The subject-matter in this portion of the work is less carefully arranged, and there is not a single refer- ence to authorities. Schmidt (op. cit. p.

368) has suggested that the peculiarities of this section are due to the use of Nicolaus of Damascus. But the remaining fragments of Nicolaus do not contain the usages in question ; traces of his style may rather ve found in the books preceding XVII. The use of a single authority for this long section is out of ‘the question, and the difference of style is probably to be accounted for by the employment of another suvepyés and amanuensis.

It is not unlikely that ‘the work was laid by for some time when the end of Book XVI. was reached. An interesting study has been made by Driiner (Untersuchungen iber gompine, Marburg, 1896, pp. 1-34) of the use made by Josephus of Thucy- ides as a model. The imitation is considerable in the earlier books of the Antiquities: from Book VI. to XII., and in XX., it is non-existent or very slight: im XIII.-XvI. it gradually increases, and reaches its climax in XVII.-xIx.

It is not confined to the diction. The narrative of incidents in the history of the Israelites has been heightened by touches from the account of the siege of Platza and the Sicilian expedition (cf. Ant. Iv. 55 with Thue. ii. 77, and Ant. tv. 91f. with Thue. vii. 83 f.) The Sicilian expedition especially has roused the Jewish historian to imitation. (See also Kennedy, Sources of NT Greek, 56f.; J. A. Ernesti, Obser- vationes Philologico-critice, etc., Leipzig, 1795).

The style of Josephus has also been influenced, though in a less degree, by a study of Herodotus (Schmidt, op. cit. 509 f.) Niese (Hist, Zeitschrift, Bd. lxxvi. 207) remarks on the language of the Jewish War that it is precious’ (yewahlt), and rich in poetical and rare words. ‘It is not the simple speech of the Atticists, but approximates to the overladen ful- ness of the Asiatic oratory.’ He finds the style of the Antiquities simpler and the poetical colour- ing almost wanting.

‘The same care, according to Niese, is not spent in the Antiquities on the avoid- ance of hiatus; in both works, however, the crasis of article and noun (e.g. rddehpod, rdcpadois, Goudriov, BJ ii. 148) appears to be the rule. vii. EDITIONS AND ‘TRANSLATIONS.—AII pre- vious editions of Josephus have been supplanted by the great critical edition of B. Niese in 7 volumes, containing a full critical apparatus and introductions on the relations of the MSS (Berlin, 1887-1895).

Niese’s only fault seems to have been a too great reliance on a single class of MSS, with the result that the true text is often to be looked for in the apparatus rather than in the text. In the manual edition of Niese, without critical ap- paratus (1888-1895), some corrections of the errors of the MSS have been introduced. On the basis of Niese’s work, Naber has constructed a recension of his own (6 vols., Teubner, 1888-1896).

Niese’s edition is indispensable to the student, but that of Naber will also be found useful as supplementing and, to some extent, improving on the work of Niese. Each of the works of Josephus has its own separate MS tradition: the MSS of the two halves of the Antiquities (I.-X., XI.-XxX.) also have their own separate histories. For this history, and for the early versions of Josephus, it will be suflicient to refer the reader to the introductions to Niese’s volumes and to Schiirer, GJV* i.

95-99. With regard to the old Latin versions it need only be stated here that we have: (1) a version of the Antiquities and the contra Apionem undertaken at the instance of Cassiodorus (de Institutione Div. Itt. 17) in the 6th cent.; (2) a version of the Jewish War commonly attributed to Rufinus ; (3) a very free Latin version of the Jewish War, which goes by the name of Hegesippus, a corruption of the name Josephus. The seven books are here compressed into five.

The original is abbreviated, freely altered, and sometimes expanded: it has the appearance of being rather a new work than a translation. It goes back to the time of Ambrose of Milan, to whose pen it has sometimes, although robably incorrectly, been attributed. — Of the ife alone no Latin version exists.—There-is a Syriac version of Book vi. of the War.

Of English translations the most serviceable, as containing the complete works, is that of Whistor, revised by Shilleto (London, 1889-1890), but the revision las been somewhat carelessly executed, and the translation is not always to be relied on. An English version of the War and the Life by Traill (London, 1862) is reported to be more reliable. Lirzraturs.—tThe literature on Josephus is immense.

For a conspectus of the more recent works, the reader must be referred to the very full bibliography given by Schiirer, op. cit. i. 100-106, to whose work the present writer is very greatly indebted. References will there be found to treatises on many interesting points, such as the chronology and geography of Josephus, which have not been touched on in the present article. H. St. J. THACKERAY. NUMBERS, HOURS, YEARS, AND DATES.*— i. Numbers and Counting.

Difficulty of fixing precise mean- ing of expressions. 1. The ‘three days’ between our Lord’s death and resurrection. 2. The erika years’ and ‘fourteen years’ of Gal 118 and 21, 8. The ‘fourteen years’ of 2 Co 122, 4. The ‘seven days’ of Ac 208, 5. The ‘twelve days’ of Ac 2441, ii. Hours of the Day. 1. Varying senses of the terms ‘ hour’ and ‘day.’ 2. ‘Hours’in the NT. The discrepancy between Mk 155 and Jn 1914, iii. Years and Dates. 1.

Dating by the years of kings and emperors, The ‘fifteenth year of Tiberius’ in Lk 31, * Cf. artt. CHRONOLOGY OF THE OT and CHRONOLOGY OF THE NT in vol. i. Most of the points dealt with in the present article concern the NT alone, although some of a wid ble laid down, particularly in the first part of § iii., will be found te apply equally to the OT, 474 NUMBERS, HOURS, AND“YEARS 2. Dating by periodically elected magisvrates. 8. Dating by priests or other officials. 4.

Devices of historians for indicating important dates. 6. Character of the dating in the NT. 6. Dating by counting from a fixed era (Seleucid, Actian, etc.) 7. The Beginning of the Year in current use. (a) According to Roman custom, year began 1 January. (6) In Asia Minor and N, Syria, year began about autumn equinox. (c) In Southern Syria, year began about spring equinox, Literature. i. NUMBERS AND COUNTING.

—Important results sometimes turn on the precise meaning of such ex- pressions as ‘six days afterwards,’ or ‘on the sixth day afterwards,’ and ‘he was ten years old,’ or ‘when he was in his tenth year.’ There is a tendency in English to differentiate between ex- pressions containing the cardinal and the ordinal numbers, so that ‘the tenth year of his age’ refers to the interval between nine and ten, while ‘ ten ate old’ means that the person in question has ived ten years and something more.

Sometimes, again, we find that, when the expression ‘six days later’ is used, the intention is not to reckon the day from which the period is counted as one of the six, whereas, when the expression ‘on the sixth day after’ is employed, the intention is to reckon the starting-point as one of the six (as, ¢.g., ‘ the sixth year after’ 1901 is 1906, but the phrase ‘six years after’ 1901 means 1907).

Generally speaking, in Greek, Roman, and Greco-Roman usage there was no such difference between the expressions with car- dinal and with ordinal numbers; but both classes of expression were used and understood as we in English tend to interpret the ordinal form. The older and popular expression in English also did not, as a rule, recognize such a difference: e.g. the idiomatic expression ‘this day eight days’ means the same day in the following week (the interval, e.g.

, from Tuesday to the following Tuesday), and ‘fifteen days’ is still sometimes used to denote an interval of a clear fortnight. The following ex- amples of ancient usage may be cited :— Cicero (ad Fam. iv. 6. 1) says that Aimilius Paullus, the conqueror of Macedonia, lost two sons within seven days. Livy (xlv. 40) tells the story in more detail, that the younger son died five days before, and the elder three days after, his triumph over Macedonia was celebrated.

Some scholars have remarked on the discrepancy be- tween these statements. But there is no discrep- ancy when the numbers are counted according to the ancient fashion. If the triumph was cele- brated, say, on the 14th day of the month, then, as Livy says, the younger son died on the 10th and the elder on the 16th; and, as Cicero says, the 16th is the seventh day after that on which the first son died. Galba adopted Piso on 10 January A.D. 69.

Then followed four complete days of sovereignty ; and on 15 January Piso, in a speech to the soldiers, spoke of the day as the sixth since his adoption (Tacitus, Hist. i. 18 and 28). There are some exceptions to this usage; but probably all could be explained as arising out of the special circumstances. Thus Tacitus else- where speaks of Piso’s reign as lasting four days (Zist. i. 19 and 48).

According to our reckoning, it lasted five clear days, from 10 to 15 January ; but there were only four unbroken days of sovereignty. The general rule that has just been stated must be applied in interpreting the numerical state- ments in the NT.— 1. The three days between the Saviour’s death and resurrection are part of Friday (viz.

the few hours that remained before sunset), the whole twenty-four hours from sunset on Friday to sunset NUMBERS, HOURS, AND YEARS on Saturday, and the few hours between sunset on Saturday and the early hour of the resurrection before sunrise on Sunday. 2. The three years and the fourteen years in Gal 1/8 21 inust be counted in the same way, the first and the last year in each period being only fractions of a year. Here the reckoning is com- plicated by the uncertainty as to how St.

Paul counted the years. Was he thinking of years of his own age; or years reckoned from the day of his conversion as prominent in his mind at the moment; or years according to the common Asia Minor and N. Syrian reckoning, with New Year in the autumn; or years according to the S. Syrian style, with New Year in the spring (like the Jewish sacred year); or years according to the Roman style, with New Year on 1 January? (see § ili. 7).

The first two of these suppositions may be at once set aside as inconsistent with the ancient custom of thought and expression: years were counted by St. Paul as beginning and endin according to the current usage, a any part the current year, however small, was counted as one year.

It would be as unreasonable to consider that he counted the years as beginning and ending according to his birthday or his conversion day as it would be to consider that he counted days as beginning and ending according to the hour of either of those events. But the real difficulty lies in determining what system of years was ordi- narily used by St. Paul in thinking and counting: in other words, what day was New Year's day in his estimation.

The present writer is not aware of any argument justifying an absolute and confident answer to this question. But the general impression made the facts stated in § ili. 7 is that St. Paul coun | according to the N. Syrian system, with the year | beginning about the autumn equinox. This gives — the general rule (stated only as probable, not as | certain), that, in reckoning the number of years | that had elapsed since any event, St.

Paul counted | the second year as beginning to run about thenext — autumn equinox: thus the interval between the event and the ensuing autumn equinox, however short, was reckoned as a year, and so with the |} interval separating the point down to which he — counts from the last preceding autumn equinox. | According to this rule, the conversion of St.

| Paul (assuming, for the moment, the traditional | day, 19 January, to be correct) and his first visit | to Jerusalem (which he says took place three years | after his conversion) might have occurred in two — successive years of the Christian era. In his way | of counting, the first year would beatan end about | 23 Sept. or 1 Oct., after the conversion, the second year would end in the autumn of the following year, and any event in Oct. or later of that year | would be in the third year.

Thus, if the con- | version were in January A.D. 31, the first visit to | Jerusalem might have occurred in Oct.-Dec. A.D, | 32, or in the first nine months of A.D. 33. On the | other hand, if St. Paul was thinking of Roman | years, the first visit could not be earlier than Jan. | of 33, and might be as late as Dec. of 33. Thusa | difference of nearly a whole year might be caused | by the slight difference between those two methods | af reckoning. 8. The statement in 2 Co 12?

is also interesting. | Fourteen years before writing, St. Paul had | enjoyed the greatest vision, and the closest com- munion with the Divine nature, that had ever — been granted him. There is probably little doubt | in the mind of almost all scholars that these words | were written during late summer or early autumn, | about six months before the last journey toJeru- | salem began.

On the scheme of chronology which | is followed in this article, this would be about Aug, NUMBERS, HOURS, AND YEARS or Sept. A.D. 56; and the year in which the vision took place would be, on the Asia Minor and N. Syrian system, the year ending in autumn A.D. 43, on the 8. Syrian system the year ending in spring A.D. 44, on the Roman system A.D. 43 (see § ili. 7). 4. In Ac 208 it is said that the deputation going to Jerusalem tarried seven daysat Troas.

As they sailed away from Troas on Monday morning, they must have arrived there on the preceding Tues- day before sunset. The journey from Philippi to Troas occupied five days, and therefore began on the Friday preceding. The five days’ journey, doubtless, included one day’s travel on Friday to Neapolis,* on Saturday they sailed for Troas, and, after a slow voyage (Ac 161), they arrived prob- ably early on Tuesday. These dates may be regarded as practically certain.

Now it seems also practically certain that St. Paul started as soon as the days of Unleavened Bread were ended, for he was eager to be in Jerusalem in time for the Feast of Pentecost. In order to reach Jerusalem he was dependent on the uncertain chance of ships ;+ he had already been in Philippi for some time, and there was no special need for him to prolong his stay for a single day after the Feast was ended.

Every consideration shows that he was bound to delay only for the festival season in Philippi, and to start immediately after. That is certainly the plain intention of the writer of Acts. The long detention in Troas, waiting for a pas- sage towards Syria, and the second shorter deten- tion in Miletus, show how uncertain was the course of a ship, and prove that St. Paul could not afford to spend any time in Philippi after the feast was ended.

On the other hand, when he had reached Czesarea, and had only a land journey along a good road, on which the rate and time could be reckoned with confidence, he was able to wait several days, and go up to Jerusalem just before Pentecost. Thus we reach the conclusion that, in the year in which St. Paul went up to Jerusalem, Passover began on a Thursday at sunset, and the days of Unleavened Bread came to an end on the following Thursday at sunset. From this it has been inferred (Ramsay, St.

Paul the Traveller, p. 289) that the journey was made in the year A.D. 57; and the discussions which have taken place on the point seem to the present writer only to have established this result more clearly.+ 5. There is much difticulty in St. Paul’s words, Ac 24" «Tt is not more than twelve days since I went up to worship at Jerusalem.’ The reckoning seems to show that it wasa little more than twelve days. St.

Paul reached Jerusalem after a journey, pre- sumably after sunset, so that, though it was only next morning that he called on St. James, yet in the reckoning both events fall in the first day. Then we seem, at first sight, to have the following list of days and events :— Ist day. Arrival after sunset ; visit to St. James next morning. 2nd ,, First day of Purification, Ac 21°.

3rd _,, Second ,, 4) 4th ,, Third ,, AS 5th ,, Fourth ,, “A * No long detention is to be expected at Neapolis, where, doubtless, ships were to be found sailing for Troas every day (see above, p. 400, also pp. 384, 389); but still a certain amount of time must have been lost there. + There were no pilgrim-ships (such as might have been got before Passover) sailing direct; and, even after a ship was found, its voyage might be broken at harbours on the way ; see above, p. 400.

: t Divergent views are stated by Mr. Turner, above, vol. i. p. 420, by Prof. Bacon in Hzpositor (1898, i. 123; 1899, ii. 351, 412; 1900, ii. 1). The latter argues on the false assumption that the strict and narrow Judaic practice of the later reaction against Roman and Christian science obtained also in the early years of the Imperial period. —— NUMBERS, HOURS, AND YEARS 475 oo 6th day. Fifth day of Purification, VAAN, es pe VOMASt hn % 8th ,, Seventh,, »”* , Ac 21%; riot; St. Paul’s speech.

9th ,, Council, Ac 22%, Dream by night, 23". 10th Conspiracy, 23. »” ‘\Journey to Antipatris begins before sunset, 23°), (J ourney to Antipatris continues by llth ,, 5 night. | Arrival in Caesarea before sunset, 23%, 12th ,, Detention in Ceesarea, 2nd day, 241, 13th ,, ” ry) ” Olds; 14th ” 9 ” ” 4th 9 ‘ 15th ae 5th ,, : trial. ” ” 99 This list seems to show that fifteen days at the least had elapsed between St.

Paul’s arrival in Jerusalem and the day when he declared that not more than twelve days had passed since he went up to Jerusalem. The explanation probably lies in Ac 217 ‘when the seven days (of purification) were about to be completed.’+ In the above list this is understood as implying that the seventh day had arrived; but it may, perhaps, be taken as merely implying ‘the seven days of purification were more than half finished, and the men were now coming near the end of the period.

’+ This seems quite consistent with the fifth day, and in that case St. Paul would be speaking on the thirteenth day since his entry into Jerusalem; and we may understand the peculiar expression ‘not more than twelve days’ as meaning ‘the thirteenth day is not yet com- pleted and past’: this form of expression shows distinct analogy with the case quoted above from Tacitus (Hist. i. 19 and 48). ii. Hours oF THE DAy.—1.

‘ Hour’ (&pa, hora) is a word used in a considerable variety of senses in the NT. The Latin hora was borrowed from the Greek (dpa), and was to a great extent deter- mined in usage by the origin. The Greek word épa meant, in a very wide and general sense, a distin- guishable period of time, a division of time marked otf by a beginning and an end, however vaguely the bounds might be indicated.

Thus épa meant, in the most general way, a measurable or estimable lapse of time; and this sense of the word never entirely disappeared, and. is found in the NT, eg. Mk 6® (twice), where it is rendered ‘day’ in both AV and RV; Mk 111, 2 Co 78, where it is rendered ‘season’in AV and RV. The dpa: rijs vuxrds and Ths huépas in Xenophon, Mem. iv. 3. 4, are not the ‘hours’ of night and of day, but the great ‘periods,’ the watches of night and the forenoon and afternoon of day.

The most characteristic division of time indi- cated by pa in early time was the season of the year; and the mythological Horaz were personitfi- cations of the Seasons. The use of the word in the sense of a division of the day, something ap- roximating to an hour in the modern usage, canity begins much before the end of the 4th cent. B.C. in the extant literature; but this quickly became the most common and widespread mean- ing of the word; and from some time, probabl early in the 3rd cent. B.

C. onwards, the Gree word in that sense was adopted in Latin. The division, which was probably of Babylonian origin * Assuming for the moment that the riot broke out on the last day of Purification ; but we shall see below that it prob- ably occurred on the fifth day. t The rendering ‘almost completed’ in AV and RV is too strong for the Greek §erAov cuvrersiobas.

t The Bezan text cuovrerouuivns d8 rHs EBdouns tyetpws, which is inconsistent with our rendering, is evidently a later alteration to secure @ more precise and definite sense t. the true Lukan text. 476 NUMBERS, HOURS, AND YEARS NUMBERS, HOURS, AND YEARS (Herod. ii.

109), was according to the duodecimal system ; and from an early time in the history of this usage traces occur both of a popular division of the period of light from sunrise to sunset (the natural day) into twelve parts or dpa, and of a scientific division of the double period of light and darkness from sunrise to sunrise, or from sunset to sunset (the civil or legal day), into twenty-four (twice twelve) parts.

Hours of the latter class, one twenty-fourth part of the fixed and unvarying period, a revolution of the earth round its axis, were of absolutely fixed and unvarying length; but the words hora, dpa, were rarely employed by the ancients in that sense: it was only astronomers that sometimes spoke of these Gpar lonuepwal, hore equinoctiales, as they were called.

In ordinary usage among the ancients, these words hora, dpa, had a different meaning, which arose out of the only means of measuring hours known and used in ordinary life by the ancients, the sun-dial. The dial, originally a very simple instrument among the Greeks, was improved, until it afforded a means of dividing the time between sunrise and sunset into twelve equal parts or hours. These hours were equal in length to each other during the same day, but varied in length from day to day.

The earliest systematic use of this division into twelve hours among the Greeks is said to have been made during the 4th sentury before Christ. While hours of this new kind were in common and popular use, the astronomers found it neces- sary for their purposes to use the equinoctial or sidereal hours of unvarying length, which they calculated by means of clepsydre or water-clocks. There often occur in the NT examples of asystem of numbering the hours of the day.

The third, the sixth, and the ninth hour, as the main divisions between the four quarters of the day, occur very often. ‘From the fifth to the tenth hour’ Ac 19% (according to the Bezan text) is a note of the hours of lecturing in a public hall of quite un- usual and even unique character in the NT ; ‘the eleventh hour’ (Mt 20°) is proverbial of the ap- proaching end of an allotted time. St.

John uses the numbers with exceptional accuracy : ‘ the tenth hour,’ 1°; ‘the seventh hour’ (in a medical obser- vation), 452, The precise meaning of these expressions in cer- tain cases has been the subject of some doubt among NT commentators; but there is absolutely no uncertainty as to the meaning in ancient pagan usage, and the doubts expressed as ea the interpretation in a few passages of Christian writ- ings are unnecessary.

A certain amount of ob- security is introduced into the subject by the use of the word ‘day’ in two different senses: the period of light trom about sunrise to sunset is called the natural day as distinguished from the period of darkness or night ; a day and a night to- gether constitute the period of the legal or civil Day.

In the following remarks we distinguish these two senses by the convention that ‘day’ means the period of light as distinguished from the night (Lichttag in German), and that ‘ Day’ means the legal period of a day and a night.

According to our own ordinary modern system of counting time, the legal Day is divided into 24 hours, and the hour is an unvarying and absolute duration of time; while the length of the day and the night are continually changing within certain limits (according to latitude), the day containing more hours and the night fewer at midsummer, and conversely at midwinter, while at the spring and the autumn equinox day and night are equal, and contain each 12 hours.

There are only the scantiest traces of such a meaning for the word ‘hour’ in ancient times, and it never occurs in popular usage, though it seems to have been known to astronomers from a very early time. The length of the ordinary ancient ‘hour’ varied continually from day to day through- out the year, The day, the period between sun rise and sunset, was divided into twelve equal parts called ‘hours’ (Jn 11%).

The division was marked by the progress of the shadow from line to line on the sun-dial; and the progress was more widely published in houses of a more preten- tious character by some such device as the blow- ing of a trumpet. In Trimalchio’s house (Pet- ronius, p.

26)* the trumpeter was an established institution ; and in the old German Imperial city of Goslar the same ancient custom was maintained by the public authorities down almost to the pres- ent time: not many years ago, and perhaps still, the trumpeter in Goslar sounded every quarter of an hour, for the division of time is carried out more minutely in modern than in ancient times.

There is hardly any trace in popular Greco- Roman usage of any definite division of time shorter than the hour: hore momento, ‘in the motion of an hour,’ z.e. the time that the shadow on the dial takes to creep from one line to the next, was a customary phrase for a brief interval (Horace, Sat. i. 1. 16). Hence the word ‘hour’ is often used in the NT to indicate a point of time, where the more emphatic expression of modern language would require some such term as ‘instant’ or ‘moment,’ e.g.

Mt 8 9%, Mk 14%, Lk 128% 4, Jn 17. But this usage may really be much more emphatic than it appears at first sight. The Latin word hora certainly, and perhaps also the Greek épa, are often used in the sense, not of the period that the shadow takes to creep from line to line on the dial, but of the brief moment in which the shadow crosses the line.

Bilfinger has beware = proved, contrary to the opinion of almost other scholars, that the latter was the moie fre- quent sense of the terms in Latin, hora prima, hora secunda: these generally indicated, not the whole time which the shadow required to move from the starting-point at sunrise to the first line, and from the first to the second line, but the moment+ when the shadow reached the first or the second line.

Hence hora sexta is frequen found, and almost always has the precise an exact sense ‘at the point of noon.’ But Bilfinger tries to push too far the view which he champions. There are certainly some cases in which hora prima means the whole period from sunrise to the moment when the shadow on the dial reaches the first division.

In truth, the ancients were far from being so accurate as modern people are; and probably many of them were hardly conscious of any difference between these two meanings, and used the term hora prima so loosely that they could hardly have specified which of the two meanings they had in mind. We ought not to try to make them out more accurate than they really were. Their vagueness in estimat- ing the divisions of time must be allowed for.

They had never been used to measure time so accurately or so minutely as we do. They had no division shorter than the hour; and they talked of the hours very loosely, making use chiefly of the prominent divisions—first, third, sixth, ninth, and twelfth hours.

The third hour meant little more than ‘during the forenoon’; and if an ordi- nary person, speaking of the third hour, were criticised and told that he should have said the sixth hour, he would probably have regarded the correction as too slight to be worth making, just as a Turkish peasant would at the present day. Modern peoples are so habituated to minute and * Cf. also Martial, viii. 67, x. 48; Juvenal, Sat. x, 215% + Bilfinger, Der burgerliche Tag.

NUMBERS, HOURS, AND YEARS accurate divisions of time, and to precise punctu- ality, that their thought and language have ac- quired a precision which is wanting in the ancient writers,* and which we must not try to force on them by strained interpretation. An example of the double meaning of the term ‘hour’ isseen if Mt 20° be compared with Plutarch, Crass. 17.

In the former passage ‘the eleventh hour’ is used metaphorically -to indicate the last oint of time, the last line and hour of the dial, ieee the allotted time, viz. the day, comes to an end, and the opportunity is lost for ever. In the latter passage ‘ the twelfth hour’ is used to convey the same proverbial sense, as being the last period, which ends when the sun sets. St.

Matthew thinks of the moment when the shadow crosses the last dividing line; Plutarch thinks of the interval that elapses between that moment and the sunset. The influence of Roman usage is here seen: dials and divisions of time seem to have been ' more familiar in the Roman time, and with greater familiarity came the use of hora to indicate the int of time when the shadow crosses the line.

When hora prima or secunda indicates a point of time, it means the moment when the shadow reaches the line at the end of the first or of the second hour after sunrise; and soon. Hence, in .

this usage, hora prima corresponds in logical sense, though not in time, to our expression ‘ one o’clock,’ The length of the hour varied, therefore, accord- ing to the length of the day: it was about 75 minutes long at midsummer, and hardly more than 45 at midwinter, while at the equinox it was exactly 60 minutes, like the hour in our modern custom. This sense of the word ‘hour’ as a period of time is found in such passages as Ac 19% 57,+ Lk 22°, Mt 20! 26”, Mk 14°”.

In popular language the varying length of the hour is alluded to in such expressions as hora estiva. In both Greek and Roman times the conception of an hour as an unvarying period of time, the twenty-fourth part of the civil or legal Day, occasionally appears in books of a more scientific character, but never in popular literature or common life.

The division of the hour into 60 minutes is said to be of Babylonian origin, and may have had an existence in scientific thought and astronomical calculations; but such a minute division played no part in popular life, never affected popular thought, and was not expressed by any word in Se language. The hour was the shortest ivision of time known to ordinary people, as has been stated above.

In ordinary usage the night was divided, not into hours but into four watches, the second of which ended at midnight.

The dial gave no means of dividing the night into hours; and the length of hours of the day could not, except with much trouble and careful adjustment taueh as only men of science would be able to give), be applied to the night by such methods of measuring as the clep- ally or water-clock, because the night hours grew shorter as the day hours grew longer, and co- incided with them in length only at the equinox.

But, by analogy from the expression hora sexta for ‘noon,’ midnight was often called hora sexta noctis (Ulpian in Digest. xli. 3.7; compare xl. 1. 1, xxviii. ]. 5; Aulus Gellius, iii. 2.11). The Greek corresponding expression is not found in the NT {probably not anywhere in strictly Greek litera- * The looseness shown by St. Luke in regard to time is noted in St. Paul the Traveller, p. 18, etc.

t In Ac 57 aipaiv rpidiv dicoryem must mean ‘the period of three hours’; it can hardly be, according to the usage described in the following paragraph, the interval that separates a line on the dial from the third following line, because that would give @ space of only two hours, according to the ancient way of counting the starting-point as the first. Lk 2259 must be inter- preted on the same analogy.

NUMBERS, HOURS, AND YEARS 477 ture), but the similar expression ‘at the third hour of night,’ implying the end of the first watch, occurs in Ac 23." In Ac 16% ‘the same hour of the night,’ the meaning is ‘instant’ or ‘ point of time,’ as in the expressions described above. It is accordingly involved in the very idea and origin of the hours in common usage that they begin from sunrise, and that the first hour ended when the shadow reached the first dividing line on the dial; and soon.

These hours are parts of the natural day, the Lichttag, and cannot be counted except as beginning with the day. The hours, as parts of the civil Day, were a totally different con- ception, which, as we have seen, never aflected or entered into popular usage and popular thought. Is it possible that those equinoctial hours might have been counted as beginning from the point when the Day was considered to begin (though only in scientific work)?

We ask, then, when the Day was considered as beginning. The legal or civil Day, comprising a complete period of day and night, was regarded as begin- ning from various points in the East and in the West. The Roman usage was the same as our modern usage: the Day was reckoned as the period from one midnight, hora sexta noctis, to the next.

In the Jewish and the Greek usage the Day was reckoned from sunset to sunset; and it is in accordance with its Eastern origin and its early development amid Greek surroundings that the Church always reckoned the ecclesiastical Day as beginning at sunset.

Bilfinger, indeed, maintains that both Greeks and Romans (except in matters of Roman law) counted the Day as beginning at daylight, either sunrise or roughly at dawn ; but Unger has conclusively refuted his arguments on this point (see his article on ‘Tagesanfang’ in Philologus, 1892, pp. 14 tf., 212 ff.), allowing only that there was a Macedonian usage (traceable at Pergamos and other places where the Macedonian calendar was used), according to which the Day was counted to begin from sunrise.

It may, how- ever, be doubted whether those seeming cases of counting the Day from sunrise may not have been simply caused by the ordinary popular custom of counting the hours of the day as beginning with the light. But however that may be, it is certain that no example has ever been quoted from the ancient writers in which the hours were counted as beginning from midnight.

Though the Roman legal Day began at midnight, yet the hours of the day were counted only as beginning from sunrise ; and the hours of the night (in the rare cases in which hours of the night were spoken of) only from sunset.

In popular usage probably no night- hours were spoken of except the third, sixth, and perhaps the ninth, as the beginnings of the second, third, and fourth watches; and those expressions were used, not because there was any device in ordinary use for dividing the night into twelve hours, but simply by analogy from the three main customary divisions of the day. 2.

There has been among some NT scholars a certain degree of hesitation about accepting as absolutely and invariably true the principle that hours were counted only as beginning from sunrise ; and some attempt has been made to show that the hours of the day were sometimes counted after a different fashion. This hesitation has been caused by the apparent discrepancy between Jn 1914 and Mk 15%.

In the former passage it is said that the trial of Jesus was concluded and the judgment on the point of being pronounced ‘about the sixth hour,’ and some more time was needed (but prob- *It is sometimes said that the Greeks had only three watches, the Romans four. This is an error arising from mis- interpretation of Pollux, i. 70. See Mr. Magan’s note om Herodotus, ix. 5, in his forthcoming edition.

478 NUMBERS, HOURS, AND YEARS ably not long) to conduct Jesus to the place of execution and raise Him on the cross, so that the Crucifixion could hardly have been consummated before 12 noon. No other estimate is given by St.

John of the lapse of time on that day, but towards sunset it was found that Jesus was dead already, though the other two sufferers were still living ; thereafter the body was taken away by Joseph and Nicodemus for burial, apparently just before the day was ended, perhaps shout 5.30 or 5.45 p.m. In the latter passage, Mk 15”, it is stated that the Crucifixion was consummated ‘at the third hour,’ 7.e. 9 a.m.

(modern time), and that at the sixth hour darkness began and lasted till the ninth hour, when Jesus died: about the beginning of evening, very soon after 6 p.m. (modern time, 1.e. sunset), Joseph took away the body for burial. It has been suggested (and the view is advocated by some high authorities) that St. John counted the hours as beginning from midnight, so that according to him judgment was pronounced on Jesus about 6 a.m.

(modern time) ; then after an interval of three hours followed the Crucifixion, and afterwards darkness began (according to the Synoptics) at noon. By this device all is shown to be in perfect harmony. It is urged that the dif- ference in the way of counting the hours was due to the fact that St. John wrote in Ephesus, and counted in this one case according to the fashion of Asia Minor as being familiar to the public for which he wrote.

An example of this supposed Asia Minor custom is sought in the martyrdom of Polycarp at the eighth hour. It is maintained that exhibitions of wild beasts, and executions by exposure to the beasts, ordinarily took place before noon (which is true),* and that therefore the eighth hour can only have been 8 a.m. (modern time). It is needless to discuss fully the case of Poly- carp;+ the facts show beyond doubt that his case was exceptional, and that he did not suffer until after noon.

He was arrested near sunset at a villa at some distance from Smyrna (to which he had retired after leaving his first refuge in a villa near the city) on a Friday, officers having been sent to arrest him in compliance with the shouts of the crowded audience at the conclusion of the venatio in the stadium at Smyrna on that day.

He was permitted to pray for two hours after arrest ; ‘when the hour for departure arrived’ (that is, evidently, early on Saturday morning) he was sonducted to Smyrna ; he was introduced for trial before the proconsul after the games in the stadium were concluded, because he could hardly have reached the city before the games began, and they would not be interrupted to allow the trial to proceed.

It is clear that the games were over for the day when the trial was held, for Philip the Asiarch (who favoured Polycarp) declared that he could not reopen them in order to comply with the demand of the crowd that Polycarp should be exposed to the beasts. The games, of course, lasted more than one day ; but it may be regarded as practically certain that they would not be con- tinued after the fifth hour.

t The interval between that hour and the eighth was occupied with the trial (for the forms of Roman law, even in a hurried trial, required some time) and the pre- paration for the execution; and the Jews, who could hardly have been present at the games on a Sabbath of especial sanctity, but who came in numbers to the trial, showed themselves active in procuring materials to burn Polycarp. Other ex- * See the full discussion in Friedlander’s Rém. Sittengesch. fii. p.

391 (349) ; also Martial, viii. 67. 4, v. 65. 8 t It is discussed at length in the Hapositor, 4th Ser. [1893], vol. vii p 220 ff. t See Hapositor, loc. cit., and the article TyRANNus, vol. iv. Martia) says ad quintam varios extendit Roma NUMBERS, HOURS, AND YEARS amples of martyrdoms which took place in the afternoon are those of Zenobius and Zenobia at Aigeai in Cilicia on a Friday at the ninth hour (Acta Sanct. 31 Oct. p. 263), and of Pionius at Smyrna at the tenth hour (Acta Pionit).

In those exceptional cases the hour was remem- bered; but in ordinary cases the execution took place early in the day, commonly forming a part of the exhibitions or-wenationes. Thus the one example that has been most confi- dently quoted to prove the existence of a peculiar way of numbering the hours in Asia Minor turns out to be an example of the ordinary custom.

In truth, the idea that in Asia Minor people counted the hours from midnight is even more improbable than it would be in other countries; for, as has been shown above, there are many instances of even the civil Day, as well as the natural day, being reckoned there to begin with sunrise. More- over, why should St. John in that one case count his hours from midnight?

It is certain and ad- mitted that elsewhere he counts them from sunrise, The more closely the subject is examined, the more clear does it 7 eobEns that the numbering of the hours in popular usage always started from the beginning of the natural day. While the other kind of hours, the equinoctial, were some- times used in scientific calculations, there is no appearance that they were numbered.

The very idea of numbering the hours is a matter of prac- tical convenience in everyday life, and has no scientific character. It must be recognized that there is an absolute —and perhaps intentional and deliberate—differ- ence between St. John and the Synoptists: the latter declare that the Crucifixion took place about three hours earlier in the day than the former admits.

With regard to this difference there arise several questions bearing on the sub- ject of this article: Which Evangelist shows him- self most attentive and observant of details of time? what is the cause of the error which must exist on one side or on the other? what is its importance? with whom does it lie? There can be no doubt that St. John is more careful about recording points and details of time.

The two disciples of the Baptist went to the place where Jesus was, ‘about the tenth hour’ (1%). Why does the historian record such a minute and in itself valueless detail? Obviously, he was natu- rally attentive to details of time, and that one remained in his memory because he had seen and known. ‘About the sixth hour’ Jesus sat down on the well of Jacob to rest (4°). The official’s son became free from the fever ‘at the seventh hour’ (452).

In these cases there is no intrinsic import- ance (as there is in the case of the Crucifixion) to make the time of day memorable. The time when they occurred is stated, because the narrator re- membered the details from personal knowledge ; and only a person attentive to time-notes would have remembered what was the hour at which each event occurred.

While the question which authority is right cannot be discussed without a far wider estimate of facts and characteristics than belongs to the present article, yet the inference from the above- stated facts is in favour of St. John’s superior trustworthiness in any statement of time; and the long interval of three hours which the Synoptists place between the consummation of the Crucifixion and the beginning of the darkness seems improb- able.

The cause of the difference in this matter lies probably or certainly in the want of attention to the lapse of time on one side or the other. St.

Jonn shows himself distinctly more attentive, through a certain personal character, whereas most of the NUMBERS, HOURS, AND YEARS NUMBERS, HOURS, AND YEARS 479 pape country people to whom the Synoptic tradition must go back are very unlikely to have paid any attention to an exact estimate of the pas- sage of time, and may quite probably have erred in their recollection of the time.

The reason for the difference lies simply in that inaccuracy in esti- mating and measuring the lapse of time which is observable in tne thought and language of the ancients. The difference in opinion is not due to correct memory being obscured in the lapse of time; it dates from the event, and would be found in accounts written at the time, if witnesses of various character had been then ordered to state their ene in writing.

As to the degree of importance to be attached co the contradiction between the witnesses, it is evident from the general considerations already stated that there is no importance in such a difference of recollection. Three events occurred successively during the day: the memory of most of the witnesses marked the sequence by record- ing that they occurred respectively at the third, the sixth, and the ninth hour, the only three divisions which the popular mind was used to note.

But those three divisions were by most people used vaguely and roughly, without any accurate estimate of the precise hour indicated on the died; and so it was in this case. The people who assigned the time of the Crucifixion as the third hour would have probably been much surprised if any one had in their presence used their testimony to pin down the event to an exact hour.

Their thought and mind were not trained to such accuracy, they saw no importance in being accurate, and they were fron: habit not capable of even attempting to be accurate in respect of the lapse of time. The sixth hour, as being midday, was better marked than any other; but even in regard to it we must allow considerable latitude when oidinary persons speak of it. St. John alone in the NT stands out as habitually careful and accurate in this respect.

The distinction between him and the other NT writers on this point is like the difference between Romans and Greeks. The Greeks made little use of the hours, and spoke little about them. The Romans used the hours in all departments of life, regulated their business and private life by them, and spoke frequently about them. St. Jolin stood on the Roman plane. iii, YEARS AND DatTes.—A convenient and practically useful system of chronology was de- veloped only very slowly in the ancient world. 1.

Dating by the years of kings and emperors.— In countries governed by monarchs, the custom of dating by the years of the current reign was natural and widespread. Such a system was of course confined almost al:solutely to the limits of the monarchy. Beyond those limits it could hardly be used, or understood, or verified. Even within those limits it had many practical disadvantages for historical purposes.

For instance, a consider- able amount of trouble was often needed to dis- cover the meaning and value of dates in past time ; as amongst ourselves it would not be obvious, without some trouble, what interval elapsed be- tween the fifteenth year of George II. and the third year of Victoria.

The difficulties of this method are of course not so serious in contem- porary dating: while a sovereign is reigning, the years of his reign, from whatever day it began, would be familiar to all the people, but after his death hardly any one remembers the exact limits of his reign. The difticulty is increased by the fact that a king reigns not for an exact number of years, but for a period that must be reckoned by years, months, and days.

If his first year is counted as running from the day of his accession, his last year is a broken one; and his successor’s reign has to be counted as beginning from a different date. Thus a literal reckoning by the exact years of each king’s reign becomes chronologically so complicated as to be extremely unwieldy and practically impossible. Some modification was therefore commonly in- troduced for chronological purposes in this method of reckoning.

The years of each king’s reign were counted according to the current and recognized system of years, and not according to the day when the reign began: either the first year of the monarch was reckoned as ending with the last day of the current year (so that this nominal first year of his reign might last in reality only a few days or months), or the last year of the former monarch was counted as running up to the end of the year in which he died, and the first year of the new monarch was counted as beginning only with the first day of the local year next after he ascended the throne, or some other device of that kind was adopted in order to facilitate chronological reckoning.

Thus either the first year of Queen Victoria would have to be reckoned as ending on 31 Dec. 1837, and her sixty-fifth year would be running at the time of her death in the beginning of 1901: but in that case the rest of the year 1901 would be reckoned by subsequent chronologists as the first ear of Edward VIl., and in historical chronology 1837 would be called as a whole the first of Victoria and 1901 the first of Edward vil. : or else the whole of 1837 would have to be reckoned to William Iv.

, 1838 would be the first year of Victoria, and 1901 would be her sixty-fourth, and would, as a whole, be reckoned as the last year of her reign. When- ever possible, it is desirable to investigate each writer’s practice from his own writings.

But, of the two alternative methods which have just been stated, the former was, on the whole, the more frequent and ordinary rule, and the one which must, in cases of doubt, be supposed to have been followed, as most likely to have been the practice of the writer in question.

On the other hand, in dating by years of the emperors, it is obvious that in Egypt the latter method was often practised, and many persons continued to date documents by the emperor who had been in office at the beginning of the year, even though his successor had already been in power for many months ; see above, p.

379° (where many of the cases alluded to are probably due, not entirely to ignorance, but to the idea that the whole current local year should be reckoned to the emperor in whose name it had begun).

Practical convenience dictated the rule, and it may be regarded as universally observed that when ordinary persons spoke of a series of years they meant, and were understood by every one to mean, the current years of the country or State to which they belonged, and not years reckoned from some arbitrary epoch, such as the birth of an individual or the accession of a sovereign.

Thus, in Egypt, the universal way of counting the ears of the Roman emperors is known to have veen according to the Egyptian year beginning on 1 Thoth (80 August). Similarly, we may be sure that in Syria people counted in ordinary usage according to the current local year (on which see § iii. 7).

There was no fixed and universal rule among chronologists regulating their practice in this respect ; and the custom of each ancient writer should as far as possible be determined separ- ately from a special study of his method. Many errors have been made by modein writers owing to misinterpretation of chronological statements according to Imperial years.

For example, in the 4580 NUMBERS, HOURS, AND YEARS NUMBERS, HOURS, AND YEARS reign of Claudius, the dates (so important for NT chronology) of the great famine (Ac 12") and of the edict expelling the Jews from Rome (Ac 18?) ave been assigned to the fourth and ninth years of his reign. Now Claudius began to reign on 25 Jan. A.D. 41, and many writers have forthwith assumed that his fourth year ran from 25 Jan. 44 to 24 Jan. 45, and his ninth from 25 Jan. 49 to 24 Jan. 50.

But that assumption is certainly wrong. We must first ask who is the authority responsible for the date, and what was his way of counting Claudius’ years. Did he follow the Roman official reckoning of years of the reign, or did he follow any of the chronologists’ methods? The authority is Orosius ; and it is clear that he followed the method which reckoned A.D. 41 as the last year of Caligula and 42 as the first year of Claudius.* Hence his authority (such as it is) places the great famine in A.

D. 45, and the edict expelling the Jews from Rome in A.D. 50. Ultimately, the value of his evidence depends entirely on the older authority or authorities on whom he was dependent:, that is a topic that has to be treated by a careful com- parative study of his account of the period as a whole. What concerns us here is that it is wrong to qrote his evidence in favour of placing those events in A.D. 44 and 49.

It lies outside of the limits of this article to in- vestigate the practice of the more scientific chrono- logists, which was not absolutely uniform. But so much is certain: the modern fashion of counting a sovereign’s years from the day of his predecessor’s death and his own accession was not followed by chronologists or historians in ancient times; and the reason lies in the hopeless cumbrousness of that method of reckoning.

Such ‘dynastic’ years, as they may be termed, were hardly thought of or reckoned by the ancients. Current years, accord- ing to local usage, alone were taken into account. The official Roman practice in reckoning the years of an emperor’s reign varied. It was cer- tainly not determined by scientific considerations of chronological convenience, and probably de- pended greatly on the choice or caprice of indi- vidual emperors.

In general, the only part of the official description or titles of the reigning emperor that gave a clue to the length of his reign was the number of times that he had held the tribunician authority,t which was apparently chosen by Augustus as the characteristic feature and the fundamental element in his tenure of authority. The real foundation of his power, of course, lay in his command of the legions. That, however, was too harsh and repellent a feature; and in B.C.

23, after years of hesitation, during which he governed as triumvir by extraordinary appointment (for periods of five years, beginning from B.C. 42), or as consul by annual election (B.C. 31-23), he finally preferred to have the tribu- nician authority as Champion of the Commons conferred on him; and henceforth in his formal list of titles the number of years during which he had held that office was stated as being equi- valent to the years of his reign.

The custom was continued by subsequent emperors. * This is pointed out by the present writer in St. Paul the Traveller (pp. 68, 254), where it is explained as due to a faulty reckoning of the years of Claudius; but in Was Christ Born at Bethlehem? p, 223, the right explanation is given that Orosius (or the older writer from whom he borrowed) reckoned inten- tionally after that fashion, We need not ask what was the New Year’s Day in Orosius’ reckoning ; it was certainly late in our vear, St.

Paul the Traveller, p. 68. + This number is always stated in the title (except in the abbreviated titles on coins): trib. pot. alone means the first year of the authority of the emperor in question. The consul- ships and the a rept salutations were also expressed numerically in his title; but the number gave no clue to the length of his reign. For example, the eleventh consulship of Augustus was in B.o, 28, but he continued to be called ‘Consul x1.’ till B.c.

5, when his twelfth consulship began. Most of the emperors of the lst cent. reckoned their years of tribunician authority from the day on which it had first been conferred on them, and disregarded the day on which their predecessor died, and on which their reign practically began. The theory was that their legal authority began when the people conferred on them tribunician and other powers, and thus made them Champion of the Commons, with the powers to make their championship effective.

The later view, which makes dynastic succession the criterion, did not rule in the Roman practice of the lst cent. and even later; and there is absolutely no justification for the common modern view, that the years of an emperor were counted in that century from the day of his predecessor’s death. According to this official Roman view— The years of Augustus began 27 June. Tiberius oF :; 7 Caligula pre March. “ » Claudius” 3, 92) Janaarye rp » Nero », 13 October (till A.D. 60).

nf »» Vespasian ,, 1 July. 7) », JLitus ’ 2” ie », Domitian ,, 13 September. Nero’s and Domitian’s days coincided with those of their predecessor's death. Claudius’ day was one day later than Caligula’s death, Caligula’s was two days later than Tiberius’ death, and the other four had no connexion whatsoever with their predecessor’s death. Nero, in A.D. 60, introduced a new way of counting his own reign, and made the change, retrospective.

His seventh year had begun on 13 October in that year, but he ordered that his eighth year should begin on 10 December A.D. 60 (for the old Republican rule was that the Tritanes of the Commons entered office always on 10 De- cember). Hence he was officially in the fifteenth year of his reign when he died on 9 June A.D. 68 (though according to our modern way of counting, by which many scholars interpret the chronological statements of the ancients, his fourteenth year was then still unfinished).

But this way of counting the years of the reign from 10 December was not imitated by any of the Ist cent. emperors except Nerva and Trajan from the year 97 onwards, and could not have had any possible influence on NT usage. a From this follows a conclusion extremely im- portant for NT chronology. There is no justifica- tion in Roman official usage for the view that when St. Luke (3!) mentioned the fifteenth year of Tiberius he was counting from the death of Augustus on 19 August A.D.

14, and meant the 12 months that began on 19 Aug. A.D. 28. We have previously seen that neither ordinary contemporary usage, nor the more scientific usage of chronolo- gists, permits such an interpretation.* We must therefore conclude that, whatever St. Luke may have meant, he certainly did not mean the year 19 August A.D. 28 to 18 August A.D. 29.

Yet the majority of modern writers assume as self-evident that that interpretation (which is founded only on modern custom and prejudice) must be what St. Luke had in mind. The question now is what bearing the special subject of this article has on the interpretation of that important date.

It may be regarded as practically certain that the custom of dating by the years of the reigning Roman emperor originated, not in Rome but in the Eastern provinces; and hence such dates are to be interpreted by Eastern, not by Roman, usage. * They would consider that his reign began on 19 August, but that his first year ended at the conclusion of the current local year (which in many places would he on 22 or 30 Sep tember).

“ 3 r ; : 4 i ped a at ail "AL NUMBERS, HOURS, AND YEARS NUMBERS, HOURS, AND YEARS 481 The Roman fashion of dating by consuls persisted in Rome far through the Imperial time; and so, e.g., the Annals of ‘Tacitus are arranged in strictly annilistic order, year by year, according to con- suls, and not by years of the emperors.

The pre- yalent method of counting in the Eastern provinces was 9 mixture of the dynastic method with the reckoning according to local years: as a general tule, the years of each emperor were counted ac- cording to the current local years, but his reign was considered to begin at the death of his pre- decessor. According to that method the fifteenth year of Tiberius would be the year beginning in ree A.D. 28, or in autumn A.D. 27, according as the local year began in spring or in autumn.

* There were, however, some exceptions to this rule about the beginning of the reign, caused chiefly by collegiate government. Thus, in Egypt, the reign of Commodus, who had been colleague of his father, M. Aurelius, for some years, was counted continuously with his father’s, as if there had been an unbroken rule from A.D. 161 to 192. Now, there were in the Ist cent. two emperors—Tiberius and Titus—who reigned for a time as colleagues of their predecessors.

In their strictly official style, both counted their years of tribunician authority from a point long anterior to their predecessors death. But Tiberius’ tribunician authority was interrupted and remained in abey- ance for a good many years, hence it was impos- sible to count the years of his reign from the first of his tribunician authority. There was, however, a second occasion when ae was assumed as colleague of his predecessor with power over all the armies in all the provinces.

This was a few days or weeks before 16 January A.D. 12, when, on his return from Pannonia, he celebrated his triumph. Before his arrival, probably at the meeting held in ordinary course on 1 January, the Senate had conferred on him those great powers; and it has been suggested with much plausibility that St. Luke (3!) considered his reign as beginning from that day, when he became col- league of Augustus, so far as the provinces and all provincial administration were concerned.

If, as many hold on other grounds to be probable, St. Luke was writing under Titus, who counted his years from his collegiate appointment, there beula be an Imperial contemporary analogy prompting the historian to this way of count- ing Tiberius’ years. One must take into account that, in the case both of Tiberius and of Augustus, it was ex- tremely difficult to tell from what date their power ought to be counted as beginning.

In both cases there were several different dates which might, with almost equal plausibility, be taken as the commencement of their reigns, while Roman custom (as we have seen) forbade that either reign should be counted as beginning from the death of the preceding ruler, the day which modern custom prefers. It is therefore impossible to arrive, on the ground of custom or etiquette, at any sure con- clusion about the sense that should be attached to the date in Lk 3!

; and the meaning can be deter- mined only by the general chronology of the life of Christ, which is not within the scope of this article. In this place there is only one further remark to make. St. Luke counted according to current years: but what, in his case, are to be understood as current years?

He was certainly influenced in various ways by Roman feeling, but it seems highly improbable that he would count according to the Roman year; and there are absolutely no other facts mentioned in his writings * It will be argued in the sequel that the local year probably begun at or soon after the autumn equinox.