Origen was in his 17th year when his father suffered martyrdom (A.D.202)  ; at eighteen he was already head of the catechetical school of Alexandria  . The Old Testament from the first engaged his attention, and, rightly judging that it could not be fruitfully studied without a knowledge of the original, he applied himself at once to the study of Hebrew.
Eus. H. E. vi.16 tosaute de eisegeto to Origenei ton theion logon apekribomene exet?sis, hos kai ten Ebraida glottan ekmathein tas te para tois Ioudaiois empheromenas prototupous autois Ebraion stoicheiois graphas ktema idion poiesasthai. Hieron. de virr. ill.54 "quis autem ignorat quod tantum in scripturis divinis habuerit studii ut etiam Hebraeam linguam contra aetatis gentisque suae naturam edisceret  ?"
The feat was perhaps without precedent, in the third century, among Christian scholars not of Jewish origin  ; in one so young it seemed prodigious to a veteran like Jerome. These studies, begun in Egypt, were continued in Palestine at Caesarea, where Origen sought shelter during the storm of persecution which burst upon Alexandria in the reign of Caracalla (A.D.216 -- 219). On his return to Egypt Origen's period of literary productivity began, and between the years 220 and 250 he gave to the world a succession of commentaries, homilies, or notes on nearly all the books of the Old Testament  . In the course of these labours, perhaps from the moment that he began to read the Old Testament in the original, he was impressed with the importance of providing the Church with materials for ascertaining the true text and meaning of the original. The method which he adopted is described by himself in his famous letter to Africanus (c. A.D.240), and more fully in his commentary on St Matthew (c. A.D.245)  .
Orig. ad Afric.5: kai tauta de phemi houchi okno tou ereunan kai tas kata Ioudaious graphas kai pasas tas hemeteras tais ekeinon sunkrinein kai horan tas en autais diaphoras, ei me phortikon goun eipein, epi polu touto (hose dunamis) pepoiekamen, gumnazontes auton ton noun en pasais tais ekdosesi kai tais diaphorais auton meta tou posos mallon askein ten hermeneian ton hebdomekonta . . . askoumen de me agnoein kai tas par ekeinois, hina pros Ioudaious dialegomenoi me prospheromen autois ta me keimena en tois antigraphois auton, kai hina sunchresometha tois pheromenois par ekeinois, ei kai en tois hemeterois ou keitai bibliois. In Matt. xv.14: ten men oun en tois antigraphois tes palaias diathekes diaphonian, theou didontos, heuromen iasasthai, kriterio chresamenoi tais loipais ekdosesin; ton gar amphiballomenon para tois o' dia ten ton antigraphon diaphonian, ten krisin poiesamenoi apo ton loipon ekdoseon, to sunadon ekeinais ephulaxamen; kai tina men obelisamen en to Ebraiko me keimena, ou tolmontes auta pante perielein, tina de met' asteriskon prosethekamen; hina delon e hoti me keimena para tois o' ek ton loipon ekdoseon sumphonos to Ebraiko prosethekamen, kai ho men boulomenos proetai auta; ho de proskoptei to toiouton, ho bouletai peri tes paradoches auton e me poiese.
2. To attempt a new version was impracticable. It may be doubted whether Origen possessed the requisite knowledge of Hebrew; it is certain that he would have regarded the task as almost impious. Writing to Africanus he defends the apocryphal additions to Daniel and other Septuagintal departures from the Hebrew text on the ground that the Alexandrian Bible had received the sanction of the Church, and that to reject its testimony would be to revolutionise her canon of the Old Testament, and to play into the hands of her Jewish adversaries athetein ta en tais ekklesiais pheromena antigrapha kai nomothetesai te adelphoteti apothesthai men tas par autois epipheromenas biblous, kolakeuein de Ioudaiois kai peithein hina metadosin hemin ton katharon). In this matter it was well, he urged, to bear in mind the precept of Prov. xxii.28, "Remove not the ancient landmark, which thy fathers have set." The same reasons prevented him from adopting any of the other versions in place of the Septuagint. On the other hand, Origen held that Christians must be taught frankly to recognise the divergences between the LXX. and the current Hebrew text, and the superiority of Aquila and the other, later versions, in so far as they were more faithful to the original; it was unfair to the Jew to quote against him passages from the LXX. which were wanting in his own Bible, and injurious to the Church herself to withhold from her anything in the Hebrew Bible which the LXX. did not represent. Acting under these convictions Origen's first step was to collect all existing Greek versions of the Old Testament. He then proceeded to transcribe the versions in parallel columns, and to indicate in the column devoted to the Septuagint the relation in which the old Alexandrian version stood to the current Hebrew text.
3. The following specimen, taken from a fragment lately discovered at Milan, will assist the reader to understand the arrangement of the columns, and to realise the general appearance of the Hexapla.
Ps. xlv. (xlvi.) 1 -- 3  .
HEBREW. HEB. TRANSLITERATED. AQUILA.
Ps. xlv. (xlvi.) 1 -- 3
SYMMACHUS. LXX. THEODOTION. 
The process as a whole is minutely described by Eusebius and Jerome, who had seen the work, and by Epiphanius, whose account is still more explicit but less trustworthy.
Eus. H. E. vi.16: tautas de hapasa [sc. tas ekdoseis] epi tauton sunagagon dielhon te pros kolon kai antiparatheis allelais meta kai autes tes Ebraion semeihoseos ta ton legomenon Hexaplon hemin antigrapha kataleloipen, idios ten Akulou kai Summachou kai Theodotionos ekdosin hama te ton hebdomekonta en tois Tetraplois epikataskeuasas. Hieron. in ep. ad Tit. iii.9: "nobis curae fuit omnes veteris legis libros quos vir doctus Adamantius in Hexapla digesserat de Caesariensi bibliotheca descriptos ex ipsis authenticis emendare, in quibus et ipsa Hebraea propriis sunt characteribus verba descripta et Graecis literis tramite expressa vicino; Aquila etiam et Symmachus, LXX. quoque et Theodotio suum ordinem tenent; nonnulli vero libri et maxime hi qui apud Hebraeos versu compositi sunt tres alias editiones additas habuit." Cf. his letter to Sunnias and Fretela (ep.106) and to Augustine (ep.112) and the preface to the Book of Chronicles. Epiph. de mens. et pond.7: tas gar hex hermeneias kai ten Ebraiken graphen Ebraikois stoicheiois kai rhemasin autois en selidi  mia suntetheikos, allen selida antiparatheton di Hellenikon men grammaton Ebraikon de lexeon pros katalepsin ton me eidoton Ebraika stoicheia . . . kai houtos tois legomenois hup' autou hexaplois e oktaplois tas men duo Ebraikas selidas kai tas hex ton hermeneuton ek parallelou antiparatheis megalen hopheleian gnoseos edoke tois philokalois. Ib.19 tas duo Ebraikas protas keimenas, meta tautas de ten tou Akula tetagmenen, meth' hen kai ten tou Summachou, epeita ten ton ob', meth' has he tou Theodotionos suntetaktai, kai hexes he pempte te kai hekte  .
It will be seen that the specimen corroborates ancient testimony in reference to the relative order of the four Greek versions (Aq., Symm., LXX., Theod.), and illustrates the method of division into corresponding kola  which made comparison easy. With regard to the order, it is clear that Origen did not mean it to be chronological. Epiphanius seeks to account for the position of the LXX. in the fifth column by the not less untenable hypothesis that Origen regarded the LXX. as the standard of accuracy (de mens. et pond.19: Origenes puthomenos ten ton ob' ekdosin akribe einai mesen tauten sunetheken, hopos tas enteuthen kai enteuthen hermeneias dielenche). As we have learned from Origen himself, the fact was the reverse; the other Greek versions were intended to check and correct the LXX. But the remark, though futile in itself, suggests a probable explanation. Aquila is placed next to the Hebrew text because his translation is the most verbally exact, and Symmachus and Theodotion follow Aquila and the LXX. respectively, because Symmachus on the whole is a revision of Aquila, and Theodotion of the LXX. As to the kola, it was of course necessary that the lines should be as short as possible when six or more columns had to be presented on each opening; and it will be seen that in the Psalms at least not more than two Hebrew words were included in a line, the corresponding Greek words being at the most three or four  . But the claims of the sense are not neglected; indeed it will appear upon inspection that the method adopted serves in a remarkable degree to accentuate the successive steps in the movement of the thought.
4. Besides the Hexapla, Origen compiled a Tetrapla, i.e. a minor edition from which he omitted the first two columns containing the Hebrew text in Hebrew and Greek characters; cf. Eus. l.c. idios ten Akulou kai Summachou kai Theodotionos ekdosin hama te ton o' en tois tetraplois epikataskeuasas  . Epiph. de mens. et pond.19 tetrapla gar eisi ta Hellenika hotan hai tou Akolou kai Summachou kai ton ob' kai Theodotionos hermeneiai suntetagmenai osi. The Tetrapla is occasionally mentioned along with the Hexapla in scholia attached to MSS. of the LXX. Thus in the Syro-Hexaplaric version at the end of Joshua it is stated that the Greek codex on which the version was based had the note: egrale ek tou hexaplou, ex hou kai paretethe; anteble?the de kai pros ton tetraploun. Cod. Q still contains two similar references to the Tetrapla (O. T. in Greek, iii., p. viii., notes). Mention is also made in the MSS. of an Octapla (cf. the Syro-Hexaplar in Job v.23, vi.28, and the Hexaplaric MSS. of the Psalter in Ps. lxxv.1, lxxxvi.5, lxxxviii.43, cxxxi.4, cxxxvi.1)  . The question arises whether the Octapla was a distinct work, or merely another name for the Hexapla in books where the columns were increased to eight by the addition of the Quinta and Sexta. Eusebius appears to support the latter view, for he speaks of the Hexapla of the Psalms as including the Quinta and Sexta (H. E. vi.16 en ge men tois hexaplois ton Psalmon meta tas etisemous tessaras ekdoseis ou monon pempten alla kai hekten kai hebdomen paratheis hermeneian). Epiphanius, on the other hand, seems to limit the Hexapla to the six columns (l. c. ton tessaron de touton selidon tais dusi tais Ebraikais sunaphtheison hexapla kaleitai; ean de kai he pempte kai he hekte hermeneia sunaphthosin . . . oktapla kaleitai. But it has been observed that when the scholia in Hexaplaric MSS. mention the Octapla they are silent as to the Hexapla, although the Octapla and the Tetrapla are mentioned together; e.g. in Ps. lxxxvi.5 we find the following note: to kata prostheken ekeito eis ten ton o' en to tetraselido (the Tetrapla), en de to oktaselido (the Octapla), , egoun dicha tou . The inference is that the name 'Octapla' sometimes superseded that of 'Hexapla' in the Psalms, because in the Psalter of the Hexapla there were two additional columns which received the Quinta and Sexta. Similarly the term 'Heptapla' was occasionally used in reference to portions of the Hexapla where a seventh column appeared, but not an eighth  . 'Pentapla' is cited by J. Curterius from cod. Q at Isa. iii.24, and Field's suspicion that Curterius had read his MS. incorrectly is not confirmed by a reference to the photograph, which exhibits en to pentaselido. Origen's work, then, existed (as Eusebius implies) in two forms: (1) the Hexapla, which contained, as a rule, six columns, but sometimes five or seven or eight, when it was more accurately denominated the Pentapla, Heptapla, or Octapla; and (2) the Tetrapla, which contained only four columns answering to the four great Greek versions, excluding the Hebrew and Greek-Hebrew texts on the one hand, and the Quinta and Sexta on the other.
5. The Hebrew text of the Hexapla was of course that which was current among Origen's Jewish teachers in the third century, and which he took to be truly representative of the original. Portions of the second column, which have been preserved, are of interest as shewing the pronunciation of the Hebrew consonants and the vocalisation which was then in use. >From the specimen already given it will be seen that k = ch, q = k, and s ,ts ,s = s and that ' h ch are without equivalent  . The divergences of the vocalisation from that which is represented by the pointing of the M. T. are more important; see Dr Taylor's remarks in D. C. B. ii. p.15 f.
In regard to Aquila, Symmachus, and Theodotion, and the minor Greek versions, Origen's task was limited to transcription under the conditions imposed by the plan of his work. But the fifth column, which contained the Hexaplaric LXX., called for the full exercise of his critical powers. If his first idea had been, as his own words almost suggest, merely to transcribe the LXX. in its proper place, without making material alterations in the text, a closer comparison of the LXX. with the current Hebrew text and the versions based upon it must soon have convinced him that this was impracticable. Let us suppose that there lay before him an Alexandrian or Palestinian MS., containing the 'common' text of the LXX. e koine, or vulgata editio , as Jerome calls it  ), i. e. the text of the Greek Bible as it was read by the Church of the third century. As the transcription proceeded, it would be seen that every column of the Greek contained clauses which were not in the Hebrew, and omitted clauses which the Hebrew contained. Further, in many places the order of the Greek would be found to depart from that of the Hebrew, the divergence being sometimes limited to a clause or a verse or two, but occasionally extending to several chapters. Lastly, in innumerable places the LXX. would be seen to yield a sense more or less at variance with the current Hebrew, either through misapprehension on the part of the translators or through a difference in the underlying text. These causes combined to render the coordination of the Alexandrian Greek with the existing Hebrew text a task of no ordinary difficulty, and the solution to which Origen was led appeared to him to be little short of an inspiration theou didontos heuromen).
Origen began by assuming (1) the purity of the Hebrew text, and (2) the corruption of the koine where it departed from the Hebrew  . The problem before him was to restore the LXX. to its original purity, i.e. to the Hebraica veritas as he understood it, and thus to put the Church in possession of an adequate Greek version of the Old Testament without disturbing its general allegiance to the time-honoured work of the Alexandrian translators. Some of the elements in this complex process were comparatively simple. (1) Differences of order were met by transposition, the Greek order making way for the Hebrew. In this manner whole sections changed places in the LXX. text of Exodus, 1 Kings, and Jeremiah; in Proverbs only, for some reason not easy to determine, the two texts were allowed to follow their respective courses, and the divergence of the Greek order from the Hebrew was indicated by certain marks  prefixed to the stichi of the LXX. column. (2) Corruptions in the koine, real or supposed, were tacitly corrected in the Hexapla, whether from better MSS. of the LXX., or from the renderings of other translators, or, in the case of proper names, by a simple adaptation of the Alexandrian Greek form to that which was found in the current Hebrew  . (3) The additions and omissions in the LXX. presented greater difficulty. Origen was unwilling to remove the former, for they belonged to the version which the Church had sanctioned, and which many Christians regarded as inspired Scripture; but he was equally unwilling to leave them without some mark of editorial disapprobation. Omissions were readily supplied from one of the other versions, namely Aquila or Theodotion; but the new matter interpolated into the LXX. needed to be carefully distinguished from the genuine work of the Alexandrian translators  . See Add. Notes.
6. Here the genius of Origen found an ally in the system of critical signs which had its origin among the older scholars of Alexandria, dating almost from the century which produced the earlier books of the LXX. The Aristarcheia semata took their name from the prince of Alexandrian grammarians, Aristarchus, who flourished in the reign of Philopator (A.D.222 -- 205, and they appear to have been first employed in connexion with his great edition of Homer  . Origen selected two of these signs known as the obelus and the asterisk, and adapted them to the use of his edition of the Septuagint. In the Homeric poems, as edited by Aristarchus, the obelus marked passages which the critic wished to censure, while the asterisk was affixed to those which seemed to him to be worthy of special attention; cf. the anecdoton printed by Gardthausen: ho de obelos pros ta athetoumena epi tou poietou egoun nenotheumena e hupobeblemena; ho de asteriskos . . . hos kalon eiremenon ton epon. Similarly, in connexion with Platonic dicta , Diogenes Laertius (platon. iii.657) used the obelus pros ten athetesin and the asterisk pros ten sumphonian ton dogmaton. As employed by Origen in the fifth column of the Hexapla, the obelus was prefixed to words or lines which were wanting in the Hebrew, and therefore, from Origen's point of view, of doubtful authority  , whilst the asterisk called attention to words or lines wanting in the LXX., but present in the Hebrew. The close of the context to which the obelus or asterisk was intended to apply was marked by another sign known as the metobelus. When the passage exceeded the length of a single line, the asterisk or obelus was repeated at the beginning of each subsequent line until the metobelus was reached.
Epiph. de mens. et pond.2, 3 ho asteriskos . . . semainei to empheromenon rhema en to Ebraiko keisthai . . . hoi de ob' hermeneutai parekan kai ouch hermeneukan . . . obelos de . . . paretithe . . . tais tes theias graphes lexesin tais para tois ob' hermeneutais keimenais, para de tois peri Akulan kai Summachon me empheromenais. Schol. ap. Tisch. not. ed cod. Sin. p.76 hosois oi obeloi proskeintai rhetois, houtoi ouk ekeinto oute para tois loipois hermeneutais oute en to Ebraiko, alla para monois tois o'; kai hosois hoi asteriskoi proskeintai rhetois, outoi en men to Ebraiko kai tois loipois hermeneutais epheronto, en de tois o' ouketi.
Occasionally Origen used asterisk and obelus together, as Aristarchus had done, to denote that the order of the Greek was at fault (anecd. ap. Gardthausen: ho de asteriskos meta obelou, hos onta men ta epe tou poietou, me kalos de keimena: schol. ap. Tisch. not. ed. Sin. l. c. pherontai men para tois o', pherontai de en to Ebraiko kai para tois loipois hermeneutais, ten thesin de monen parallassousin hoi loipoi kai to Ebraikon para tous o'; hothen obelistai en tauto kai esteristai, hos para pasi men pheromena, ouk en tois autois de topois: also ap. mon. sacr. ined. iii. p. xvii. ta de esterismena en tauto kai obelismena rheta . . . hos para pasi men pheromena, ouk en tois autois de topois). The Aristarchian (or as they are usually called by students of the Old Testament, the Hexaplaric) signs are also used by Origen when he attempts to place before the reader of his LXX. column an exact version of the Hebrew without displacing the LXX. rendering. Where the LXX. and the current Hebrew are hopelessly at issue, he occasionally gives two versions, that of one of the later translators distinguished by an asterisk, and that of the LXX. under an obelus  .
The form of the asterisk, obelus, and metobelus varies slightly. The first consists of the letter x, usually surrounded by four dots ( , the chi teriestigmenon); the form occurs but seldom, and only, as it seems, in the Syro-Hexaplar. The orelos, 'spit' or 'spear,' is represented in Epiphanius by , but in the MSS. of the LXX. a horizontal straight line ( -- )  has taken the place of the original form, with or without occupying dot or dots ( ); the form was known as a lemniscus, and the form as a hypolemniscus. Epiphanius indeed (op. cit., c.8) fancies that each dot represents a pair of translators, so that the lemniscus means that the word or clause which the LXX. adds to the Hebrew had the support of two out of the thirty-six pairs which composed the whole body, whilst the hypolemniscus claims for it the support of only one pair. This explanation, it is scarcely necessary to say, is as baseless as the fiction of the cells on which, in the later Epiphanian form, it rests. Other attempts to assign distinct values to the various forms of the obelus have been shewn by Field to be untenable  . The metobelus is usually represented by two dots arranged perpendicularly (:), like a colon; other forms are a sloping line with a dot before it or on either side (/., ·/.), and in the Syro-Hexaplar and other Syriac versions a mallet . The latter form, as the least ambiguous, is used in Field's great edition of the Hexapla, and in the apparatus which is printed under the text of the LXX. version of Daniel in the Cambridge manual Septuagint.
Certain other signs found in Hexaplaric MSS. are mentioned in the following scholion Euagriou sch., one of the scholia eis tas paroimias printed in the Notitia ed. cod. Sin., p.76, from a Patmos MS.; see Robinson, Philocalia, pp. xiii., xvii. ff.): eisin  hosa protetagmenon echousi ton arithmon hode; hosa Origenen epigegrammenon echei touto to monosullabo, . . . hosa de peri diaphonias rheton tinon ton en to edaphio e ekdoseon estin scholia, haper kai kato neneukuian periestigmenen echei protetagmenen, ton antibeblekoton to biblion estin; hosa de amphibolos exo keimena rheta exo neneukuian periestigmenen echei protetagmenen, dia ta scholia prosetethesan kat' auta tou megalou eirekotos didaskalou, hina me doxe kata kenou to scholion pheresthai, en pollois men ton antigraphon ton rheton houtos echonton, en touto de me houtos keimenon e med' holos pheromenon, kai dia touto prostethenton.
The following extract from the great Hexaplaric MS. known as G will enable the student, to whom the subject may be new, to practise himself in the interpretation of the signs. He will find it instructive to compare the extract with his Hebrew Bible on the one hand and the text of Cod. B (printed in the Cambridge LXX.) on the other  .
Joshua xi.10 -- 14 (Cod. Sarravianus).
kai epestrepsen is en to kairo ekeino katelabeto ten : asor kai ton basilea autes apekteinen en rom phaia : en de asor po proteron archousa paso ton basileion touton kai apekteina pan enpneon o : en aute en stomati xiphous kai exolethreusan : -- pantas : kai ou kateliphthe en aute enpneon kai ten asor enepresen en puri kai pasas tas poleis ton basileion touto : kai pantas : tous basileis auton elaben is kai aneilen autous en stomati xiphous exolethreusen autous on tropon sunetaxe Moses o pais ku; alla pasas tas poleis tas kechomatismenas auton : ouk enepresen inl plen ten : asor monen auten : enepresen is kai pata ta skula autes ta ktene : epronomeusan eautois oi inl kata to rema ku o ene teilato to iu : autous de pantas exolethreusen en stomati xiphous eos apolesen autous ou katilipon auto : oude en enpneon * * *
7. The Hexapla was completed, as we have seen, by A.D.240 or 245; the Tetrapla, which was a copy of four columns of the Hexapla, followed, perhaps during Origen's last years at Tyre.  A large part of the labour of transcription may have been borne by the copyists who were in constant attendance on the great scholar, but he was doubtless his own diorthotes, and the two Hebrew columns and the LXX. column of the Hexapla were probably written by his own hand.
Eusebius in a well-known passage describes the costly and laborious process by which Origen's commentaries on Scripture were given to the world: H. E. vi.23 tachugraphoi gar auto pleious e hepta ton arithmon paresan hupagoreuonti, chronois tetagmenois allelous ameibontes, bibliographoi te ouch hettous hama kai korais epi to kalligraphein eskemenais; hon hapanton ten deousan ton epitedeion aphthonon periousian ho Ambrosios parestesato. Two of these classes of workers, the bibliographoi and kalligraphoi (cf. Gardthausen, Gr. Palaeographie, p.297), must have found ample employment in the preparation of the Hexapla. The material used was possibly papyrus. Although there are extant fragments of writing on vellum which may be attributed to the second century, "there is every reason to suppose that to the end of the third century papyrus held its own, at any rate in Egypt, as the material on which literary works were written" (Kenyon, Palaeography of Gk papyri, p.113 f.; on the size of existing papyrus rolls, see p.16 ff.). This view receives some confirmation from Jerome's statement (ep.141) that Acacius and Evagrius endeavoured to replace with copies on parchment some of the books in the library at Caesarea which were in a damaged condition ("bibliothecam . . . ex parte corruptam . . . in membranis instaurare conati sunt")  . According to Tischendorf (prolegg. in cod. Frid. Aug. § 1) cod. ' was written on skins of antelopes, each of which supplied only two leaves of the MS. The Hexapla, if copied in so costly a way, would have taxed the resources even of Origen's generous ergodioktes.
It is difficult to conceive of a codex or series of codices so gigantic as the Hexapla. Like the great Vatican MS., it would have exhibited at each opening at least six columns, and in certain books; like the Sinaitic MS., eight. Its bulk, even when allowance has been made for the absence in it of the uncanonical books, would have been nearly five times as great as that of the Vatican or the Sinaitic Old Testament. The Vatican MS. contains 759 leaves, of which 617 belong to the Old Testament; when complete, the O. T. must have occupied 650 leaves, more or less. From these data it may be roughly calculated that the Hexapla, if written in the form of a codex, would have filled 3250 leaves or 6500 pages  ; and these figures are exclusive of the Quinta and Sexta, which may have swelled the total considerably. Even the Tetrapla would have exceeded 2000 leaves. So immense a work must have been the despair of copyists, and it is improbable that any attempt was made to reproduce either of the editions as a whole. The originals, however, were long preserved at Caesarea in Palestine, where they were deposited, perhaps by Origen himself, in the library of Pamphilus. There they were studied by Jerome in the fourth century (in Psalmos comm. ed. Morin., p.5: "hexaplous Origenis in Caesariensi bibliotheca relegens"; ib. p.12: "cum vetustum Origenis hexaplum psalterium revolverem, quod ipsius manu fuerat emendatum"; in ep. ad Tit.: "nobis curae fuit omnes veteris legis libros quos v. d. Adamantius in Hexapla digesserat de Caesariensi bibliotheca descriptos ex ipsis authenticis emendare." There also they were consulted by the writers and owners of Biblical MSS.; compare the interesting note attached by a hand of the seventh century to the book of Esther in cod. ' : anteblethe pros palaiotaton lian antigraphon dediorthomenon cheiri tou hagiou `arturos Pamphilou; pros de to telei tou autou palaiotatou bibliou . . . huposemeiosis tou autou marturos hupekeito echousa houtos; (O. T. in Greek, ii. p.780); and the notes prefixed to Isaiah and Ezekiel in Cod. Marchalianus (Q); the second of these notes claims that the copy from which Ezekiel was transcribed bore the subscription , (ib. iii. p. viii.)  . The library of Pamphilus was in existence in the 6th century, for Montfaucon (biblioth. Coisl. p.262) quotes from Coisl.202  , a MS. of that century, a colophon which runs: anteblethe de he biblos pros to en Kaisaria antigraphon tes bibliothekes tou hagiou Pamphilou cheiri gegrammenon autou. But in 638 Caesarea fell into the hands of the Saracens, and from that time the Library was heard of no more. Even if not destroyed at the moment, it is probable that every vestige of the collection perished during the vicissitudes through which the town passed between the 7th century and the 12th  . Had the Hexapla been buried in Egypt, she might have preserved it in her sands; it can scarcely be hoped that the sea-washed and storm-beaten ruins of Kaisariyeh cover a single leaf.
LITERATURE. Fragments of the Hexapla were printed by Peter Morinus in his notes to the Roman edition of the Septuagint (1587). Separate collections have since been published by J. Drusius (Vet. interpretum Graecorum . . . fragmenta collecta . . . a Jo. Drusio, Arnheim, 1622), Bernard Montfaucon (Origenis Hexaplorum quae supersunt, Paris, 1713), and F. Field (Oxford, 1875), whose work has superseded all earlier attempts to recover the Hexapla. A fuller list may be seen in Fabricius-Harles, iii.701 ff. Materials for an enlarged edition of Field are already beginning to accumulate; such may be found in Pitra, Analecta sacra, iii. (Venice, 1883), p 551 ff.; E. Klostermann, Analecta zur . . . Hexapla (Leipzig, 1895), G. Morin, Anecdota Maredsolana iii.1 (Mareds., 1895; cf. Expositor, June 1895, p.424 ff.), and the Oxford Concordance. Among helps to the study of the Hexapla, besides the introductions already specified, the following may be mentioned: the Prolegomena in Field's Hexapla, the art. Hexapla in D. C. B. by Dr C. Taylor; the introduction to Dr Driver's Notes on Samuel (p. xliii. ff.), and Harnack-Preuschen, Gesch. altchristt. Litt. i. p.339 ff. For the literature of the Syro-Hexaplaric version see c. iv.
8. The Hexapla as a whole was perhaps too vast to be copied  , and copies even of particular books were rarely attempted; yet there was nothing to forbid the separate publication of the fifth column, which contained the revised Septuagint. This idea presented itself to Pamphilus and his friend Eusebius, and the result was the wide circulation in Palestine during the fourth century of the Hexaplaric LXX., detached from the Hebrew text and the other Greek versions, but retaining, more or less exactly, the corrections and additions adopted by Origen with the accompanying Hexaplaric signs. "Provinciae Palestinae," writes Jerome in his preface to Chronicles, "codices legunt quos ab Origene elaboratos Eusebius et Pamphilus vulgaverunt." Elsewhere  he warns his correspondents "aliam esse editionem quam Origenes et Caesariensis Eusebius omnesque Graeciae tractatores koinen (id est communem) appellant atque vulgatam . . ., aliam LXX. interpretum quae in exaplois codicibus reperitur . . et Ierosolymae atque in orientis ecclesia decantatur." The Hexaplaric text receives his unhesitating support: "ea autem quae habetur in hexaplois . . . ipsa est quae in eruditorum libris incorrupta et immaculata LXX. interpretum translatio reservatur  ." This edition, sometimes described as to Eusebiou or to Palaistinaion, or simply Or[igenes], is mentioned with great respect in the scholia of MSS. which do not on the whole follow its text. Specimens of such notes have already been given; they usually quote the words in which Pamphilus describes the part borne by himself and his friends respectively in the production of the book. Thus a note quoted by an early hand in cod. ' at the end of 2 Esdras says, Antoninos antebalen, Pamphilos diorthosa. The subscription to Esther ends Antoninos homologetes antebalen, Pamphilos diorthosato [to] teuchos en te phulake. The scholion prefixed to Ezekiel in Q introduces the name of Eusebius, assigning him another function: Eusebios ego ta scholia paretheka; Pamphilos kai Eusebios diorthosanto. In its subscription to 1 Kings the Syro-Hexaplar quotes a note which runs: Eusebios diorthosamen hos akribos edunamen. It would seem as though the work of comparing the copy with the original was committed to the otherwise unknown  Antoninus, whilst the more responsible task of making corrections was reserved for Pamphilus and Eusebius  . Part of the work at least was done while Pamphilus lay in prison, i.e. between A.D.307 and 309, but it was probably continued and completed by Eusebius after the martyr's death.
The separate publication of the Hexaplaric LXX. was undertaken in absolute good faith; Pamphilus and Eusebius believed (as did even Jerome nearly a century afterwards) that Origen had succeeded in restoring the old Greek version to its primitive purity, and they were moved by the desire to communicate this treasure to the whole Church. It was impossible for them to foresee that the actual result of their labours would be to create a recension of the LXX. which was a mischievous mixture of the Alexandrian version with the versions of Aquila and Theodotion. The Hexaplaric signs, intended for the use of scholars, lost their meaning when copied into a text which was no longer confronted with the Hebrew or the later versions based upon it; and there was a natural tendency on the part of scribes to omit them, when their purpose was no longer manifest.
When we consider that the Hexaplaric Septuagint claimed to be the work of Origen, and was issued under the authority of the martyr Pamphilus and the yet greater Bishop of Caesarea, we can but wonder that its circulation was generally limited to Palestine  . Not one of our uncial Bibles gives the Hexaplaric text as a whole, and it is presented in a relatively pure form by very few MSS., the uncials G and M, which contain only the Pentateuch and some of the historical books, and the cursives 86 and 88 (Holmes and Parsons), which contain the Prophets. But a considerable number of so-called Hexaplaric codices exist, from which it is possible to collect fragments not only of the fifth column, but of all the Greek columns of the Hexapla; and a still larger number of our MSS. offer a mixed text in which the influence of the Hexaplaric LXX., or of the edition published by Pamphilus and Eusebius, has been more or less extensively at work  . The problems presented by this and other causes of mixture will come under consideration in the later chapters of this book.
9. While the Hexaplaric Septuagint was being copied at Caesarea for the use of Palestine, Hesychius was engaged in correcting the common Egyptian text.
Hieron. in praef. ad Paralipp.: "Alexandria et Aegyptus in Septuaginta suis Hesychium laudat auctorem"; cf. adv. Rufin. ii. where the statement is repeated  , and praef. in Evangelia, where the revision of Hesychius is represented as having included both Testaments, and his O. T. work is condemned as infelicitous ("nec in V.T. post LXX. interpretes emendare quod licuit"); the Hesychian revision of the Gospels is censured by the Decretum Gelasii, which even denounces them as apocryphal ("evangelia quae falsavit Hesychius, apocrypha").
It is not easy to ascertain who this Hesychius was. The most conspicuous person of that name is the lexicographer, and he has been identified with the reviser of the Greek Bible  . But later researches shew that Hesychius the lexicographer was a pagan who lived in the second half of the fourth century. The author of the Egyptian revision was more probably  the martyr Bishop who is mentioned by Eusebius in connexion with Phileas Bishop of Thmuis, Pachymius, and Theodorus (H. E. viii.13 Phileas te kai Hesuchios kai Pachumios kai Theodoros ton amphi ten Aigupton ekklesion episkopoi). The four names appear together again in a letter addressed to Meletius (Routh, rell. sacr. iv. p.91 ff.); and Eusebius has preserved a pastoral written by Phileas in prison in view of his approaching martyrdom (H. E. viii.10). Phileas was a distinguished scholar (H. E. viii.9 diaprepsas . . en . . tois kata philosophian logois, ib.10 ton exothen mathematon heneka pollou logou axion . . . tou hos alethos philosophou . . marturos), and the association of his name with that of Hesychius suggests that he may have shared in the work of Biblical revision. It is pleasant to think of the two episcopal confessors employing their enforced leisure in their Egyptian prison by revising the Scriptures for the use of their flocks, nearly at the same time that Pamphilus and Eusebius and Antoninus were working under similar conditions at Caesarea. It is easy to account for the acceptance of the Hesychian revision at Alexandria and in Egypt generally, if it was produced under such circumstances.
To what extent the Hesychian recession of the Old Testament is still accessible in MSS. and versions of the LXX. is uncertain. As far back as 1786 Münter threw out the very natural suggestion that the Egyptian recession might be found in the Egyptian versions. In his great monograph on the Codex Marchalianus Ceriani takes note that in the Prophets, with the exception perhaps of Ezekiel, the original text of that great Egyptian MS. agrees closely with the text presupposed by the Egyptian versions and in the works of Cyril of Alexandria, and that it is supported by the cursive MSS.26, 106, 198, 306; other cursives of the same type are mentioned by Cornill  as yielding an Hesychian text in Ezekiel. For the remaining books of the LXX. we have as yet no published list of MSS. containing a probably Hesychian text, but the investigations now being pursued by the editors of the larger Cambridge LXX. may be expected to yield important help in this direction  .
10. Meanwhile the rising school of Antioch was not inactive in the field of Biblical revision. An Antiochian recession of the koine had in Jerome's time come to be known by the name of its supposed author, the martyr Lucian  .
Hieron. praef. in Paralipp.: "Constantinopolis usque Antiochiam Luciani martyris exemplaria probat." Cf. (Ep. cvi.) ad Sunn. et Fret.2 "[he koine] . . . a plerisque nunc Loukianos dicitur." Ps.-Athan. syn. sacr. script. hebdome palin kai teleueaia hermeneia tou hagiou Loukianou tou megalou asketou kai marturos, hostis kai autos tais progegrammenais ekdosesi kai tois Ebraikois entuchon kai epopteusas met' akribeias ta leiponta e kai peritta tes aletheias rhemata kai diorthosamenos en tois oikeiois ton graphon topois exedoto tois christianois adelphois; hetis de kai hermeneia meta ten athlesin kai marturian tou autou hagiou Loukianou ten gegonuian epi Diokletianou kai Maximianou ton turannon, egoun to idiocheiron hautou tes ekdoseos biblion, heurethe en Nikomedeia epi Konstantinou basileos tou megalou para Ioudaiois en toicho purgisko perikechrismeno koniamati eis diaphulaxin (cf. the Acts of Lucian in Bolland. i. p.363). Suidas s.v. houtos tas hieras biblous theasamenos polu to nothon eisdexamenas, tou ge chronou lumenamenou polla ton en autais kai tes sunechous aph' heteron eis hetera metatheseos . . . autos hapasas analabon ek tes Ebraidos epaneneosato glosses. Cf. also Cyr. Alex. in Psalmos praef.
Lucian, who was born at Samosata, began his studies at Edessa, whence he passed to Antioch at a time when Malchion was master of the Greek School (Eus. H. E. vii.29, Hieron. de virr. ill.71). At Antioch Lucian acquired a great reputation for Biblical learning (Eus. H. E. ix.6 tois hierois mathhumasi sunkekrotemenos, Suid. s.v. auten [sc. ten Ebraida glossan] hos ta malista en ekribokos). From some cause not clearly explained Lucian was under a cloud for several years between A.D.270 and 299 (Theodoret  , H. E. i.3 aposunagogos emeine trion episkopon poluetous chronou). On his restoration to communion he was associated with Dorotheus, who was a Hebrew scholar, as well as a student of Greek literature (Eus. H. E. vii.32 philokalos d' houtos peri ta theia grammata kai tes Ebraion epemelethe glottes, hos kai autais tais Ebraikais graphais epistemonos entunchanein; en de houtos ton malista eleutherion, propaideias te tes kath' Hellenas ouk amoiros). As Pamphilus was assisted by Eusebius, as Phileas and others were probably associated with Hesychius, so (the conjecture may be hazarded) Dorotheus and Lucian worked together at the Antiochian revision of the Greek Bible. If, as Dr Hort thought, "of known names Lucian's has a better claim than any other to be associated with the early Syrian revision of the New Testament  ," the Syrian revision of the Old Testament, which called for a knowledge of Hebrew, may have been due more especially to the Hebraist Dorotheus. Lucian, however, has the exclusive credit of the latter, and possibly was the originator of the entire work. If we may believe certain later writers, his revision of the LXX. was on a great scale, and equivalent to a new version of the Hebrew Bible; Pseudo-Athanasius goes so far as to call it the hebdome hermeneia, placing it on a level with the Greek versions of the Hexapla. But Jerome's identification of 'Lucian' with the koine presents quite another view of its character and one which is probably nearer to the truth. It was doubtless an attempt to revise the koine in accordance with the principles of criticism which were accepted at Antioch. In the New Testament (to use the words of Dr Hort  ) "the qualities which the authors of the Syrian text seem to have most desired to impress on it are lucidity and completeness . . . both in matter and in diction the Syrian text is conspicuously a full text." If the Lucianic revision of the LXX. was made under the influences which guided the Antiochian revision of the New Testament, we may expect to find the same general principles at work  , modified to some extent by the relation of the LXX., to a Hebrew original, and by the circumstance that the Hebrew text current in Syria in the third century A.D. differed considerably from the text which lay before the Alexandrian translators.
We are not left entirely to conjectures. During his work upon the Hexapla  Field noticed that in an epistle prefixed to the Arabic Syro-Hexaplar  , the marginal letter (L) was said to indicate Lucianic readings. Turning to the Syro-Hexaplar itself, he found this letter in the margin of 2 Kings (= 4 Regn.) at cc. ix.9, 28, x.24, 25, xi.1, xxiii.33, 35, But the readings thus marked as Lucianic occur also in the cursive Greek MSS.19, 82, 93, 108; and further examination shewed that these four MSS. in the Books of Kings, Chronicles, and Ezra-Nehemiah agree with the text of the LXX. offered by the Antiochian fathers Chrysostom and Theodoret, who might have been expected to cite from 'Lucian.' Similar reasoning led Field to regard codd.22, 36, 48, 51, 62, 90, 93, 144, 147, 233, 308 as presenting a more or less Lucianic text in the Prophets. Meanwhile, Lagarde had independently  reached nearly the same result, so far as regards the historical books. He satisfied himself that codd.19, 82, 93, 108, 118  , had sprung from a common archetype, the text of which was practically identical with that of the LXX. as quoted by Chrysostom, i.e., with the Antiochian text of the fourth century, which presumably was Lucianic. Lagarde proceeded to construct from these and other sources a provisional text of Lucian, but his lamented death intercepted the work, and only the first volume of his Lucianic LXX. has appeared (Genesis -- 2 Esdr., Esther).
The following specimen will serve to shew the character of Lucian's revision, as edited by Lagarde; an apparatus is added which exhibits the readings of codd. B and A.
3 Regn. xviii.22-28.
^22 kai eipen Helias pros ton laon Ego hupoleleimmai prophetes kuriou prophetes monotatos, kai hoi prophetai tou Baal tetrakosioi kai pentekonta andres, kai hoi prophetai ton alson tetrakosioi. ^23 dotosan oun hemin duo boas, kai eklexasthosan heautois ton hena kai melisatosan kai epithetosan epi xula kai pur me epithetosan; kai ego poieso ton boun ton allon, kai pur ou me epitho. ^24 kai boate en onomati theon humon, kai ego epikalesomai en onomati kuriou tou theou mou, kai estai ho theos hos an epakouse semeron en puri, houtos esti theos. kai apekrithe pas ho laos kai eipen Agathos ho logos hon elalesas. ^25 kai eipen Helias tois prophetais tes aischunes Eklexasthe heautois ton boun ton hena , hoti humeis polloi, kai poiesate protoi, kai epikaleisthe en onomati theon humon, kai pur me epithete. ^26 kai elabon ton boun kai epoiesan, kai epekalounto en onomati tou Baal kai eipon Epakouson hemon, ho Baal, epakouson hemon. kai ouk en phone kai ouk en akroasis. kai dietrechon epi tou thusiasteriou hou epoiesan. ^27 kai egeneto mesembria, kai emukterisen autous Helias ho Thesbites kai prosetheto legon Epikaleisthe en phone megale hama, mepote adoleschia tis estin auto, kai hama mepote chrematizei autos e mepote katheudei, kai exanastesetai. ^28 kai epekalounto en phone megale kai katetemnonto kata ton ethismon auton en machairais kai en seiromastais heos ekchuseos haimatos ep' autous.
22 Eleiou BA kuriou] pr tou BA om prophetes 2^0 BA oi prophetai 2^0] om oi A tou alsous BA om tetrakosioi 2^0 A 23 om oun BA om kai epith. epi xula A xula] ton xulon B ton allon] + kai doso epi ta xula A 24 theon] theou A ean BA om semeron BA om esti BA apekrithesan BA eipon B eipan A agathos o logos on] kalon to rema o BA 25 Eleiou BA boun] moschon BA kai poi. protoi oti polloi umeis BA epikalesasthe B theon] theou BA 26 elaben A boun] moschon BA + on edoken autois A Baal 1^0] oti BA tis estin auto] auto estin BA katheudei] + autos BA 28 kata ton ethismon auton] om B kata to krima auton A machaira B om en 3^0 B
A comparison of 'Lucian' in this passage with the two great uncials of the LXX. reveals two classes of variants in the former. (1) Some of the changes appear to be due to a desire to render the version smoother or fuller, e.g. Helias for Heleiou, the repetition of prophetes before monotatos, the substitution of ton alson for tou alsous, of apekrithe for apekrithesan, and of agathos ho logos for kalon to rhema, and the addition of semeron. (2) Others seem to indicate an attempt to get nearer to the Hebrew, e.g. dotosan oun (vytgv), boun (pr); or an adherence to an older reading which the Hexaplaric LXX. had set aside, e.g. the omission of hon edoken autois  and ek proithen heos mesembrias. On the other hand Lucian follows the current Hebrew in kata ton ethismon auton, though he substitutes the easier ethismos for Aquila's krima, which cod. A has taken over from the Hexapla.
Professor Driver, as the result of a wider examination, points out  that the Lucianic recession is distinguished by (1) the substitution of synonyms for the words employed by the LXX.; (2) the occurrence of double renderings; (3) the occurrence of renderings "which presuppose a Hebrew original self-evidently superior in the passages concerned to the existing Massoretic text." The last of these peculiarities renders it of great importance for the criticism of the Hebrew Bible.
Lucian suffered martyrdom at Nicomedia under Maximin in the year 311 or 312  . According to the Pseudo-Athanasian Synopsis, his recension of the LXX. was subsequently discovered at Nicomedia, bricked up in a wall. The story may have arisen from a desire to invest the hebdome (as 'Lucian' is called by the author of the Synopsis) with the same air of romance that belonged to the Quinta and Sexta, both of which were found, as he asserts, en pithois. It is more probable that copies were circulated from Antioch in the ordinary way, and that some of these after the persecution reached Nicomedia and Constantinople. The name of Lucian would be enough to guarantee the general acceptance of the work. He died in the peace of the Church, and a martyr; on the other hand his name was in high repute with the Arian leaders, who boasted of being sulloukianistai  . Moreover, a revision which emanated from Antioch, the "ecclesiastical parent of Constantinople  ," would naturally take root in the soil of the Greek East. In all dioceses which felt the influences of those two great sees, the Lucianic LXX. doubtless furnished during the fourth and fifth centuries the prevalent text of the Greek Old Testament  .
11. The result of these multiplied labours of Christian scholars upon the text of the LXX. was not altogether satisfactory. Before the time of Jerome much of the original text of the Alexandrian Bible had disappeared. Men read their Old Testament in the recension of Lucian, if they lived in North Syria, Asia Minor, or Greece; in that of Hesychius, if they belonged to the Delta or the valley of the Nile; in Origen's Hexaplaric edition, if they were residents at Jerusalem or Caesarea. Thus, as the scholar of Bethlehem complains, the Christian world was divided between three opposing texts ("totus . . . orbis hac inter se trifaria varietate compugnat  "). To Jerome, as a Palestinian and an admirer of Origen's critical principles, the remedy was simple; the Hexaplaric text, which had been assimilated to the Hebraica veritas, ought everywhere to take the place of the koine represented by Hesychius or Lucian. Fortunately the task was beyond his strength, and MSS. and versions still survive which represent more or less fully the three recessions of the fourth century. But the trifaria varietas did not continue to perplex the Church; a fusion of texts arose which affected the greater part of the copies in varying proportions. No one of the rival recessions became dominant and traditional, as in the case of the New Testament  ; among the later MSS, groups may be discerned which answer more or less certainly to this recession or to that, but the greater number of the cursives present a text which appears to be the result of mixture rather than of any conscious attempt to decide between the contending types.
 Eus. H. E. vi. 2.  Hieron. de virr. ill. 54.  Cf. ep. ad Paulam.  See D. C. B. art. Hebrew Learning (ii. p. 351 ff.).  See D. C. B. art. Origenes, iv. p. 129 ff.  Cf. Bp Westcott in D. C. B, iv. p. 99: "it was during this period (i.e. before A.D. 215) in all probability that he formed and partly executed his plan of a comparative view of the LXX. in connexion with the other Greek versions."  Cf. Un palimpsesto Ambrosiano dei Salmi Esapli (Giov. Mercati) in Atti d. R. Accademia d. Scienze di Torino, 10 Apr. 1896; and E. Klostermann, die Mailänder Fragmente der Hexapla. The MS. does not supply the Hebrew column.  In the MS. lanou appears in the third column, where it has displaced Aquila's rendering.  MS. eurethes.  Or Quinta? Cf. H. Lietzmann in G. G. A. 1902, v., p. 332: "die letzte Columne ist nicht, wie man anfangs glaubte, Theodotion, sondern die Quinta mit Interlinearvarianten."  With marginal variants, eis to telos, psalmos (LXX.).  With interlinear variant tois huiois (Th.).  With marginal variants, eis to telos, psalmos (LXX.).  MS. 1 a manu hemin (? Aq. Sym.).  With interlinear variant heurethesetai hemin.  With interlinear variant tais heurousais hemas (LXX.).  MS. tais.  With interlinear variant metatithesthai (LXX.).  On selis, cf. Sir E. Maunde Thompson, Handbook of Greek and Latin Palaeography, p, 58.  See also ib. 18 sq.; Hieron. Praef. in Paral., and in ep. ad Tit., c. iii.  Used here loosely as = kommata, the kolon being properly a line consisting of a complete clause, and of 8--17 syllables: cf. E. M. Thompson, Gk and Lat. Palaeography, p. 81 f.; J. R. Harris, Stichometry, p. 23 f.  In the earlier Cairo palimpsest even such words as 'l and me had each a line to itself; see Nestle in Hastings' D.B. iv. 443.  Epikataskeuazein is insuper vel postea concinnare (Field, prolegg. p. xii.); cf. Dio Cass. l. 23 ta skaphe kateskeuase . . . kai ep' auta purgous epekateskeuase. Oeconomus (iv. 873), who regards the Tetrapla as the earlier work, understands Eusebius to mean only that Origen added to the LXX. the three columns containing A'S'Th' .  Field, Hexapla, ii. ad loc.; cf. Hieron. in Psalmos (ed. Morin.), p. 66.  It occurs (e.g.) in the Hexaplaric Syriac at 2 Kings 16:2.  Cf. the practice of Aquila (Burkitt, Fragments of the Books of Kings acc. to Aquila, p. 14).  Ep. ad Sunn. et Fret.  See Driver, Samuel, p. xlvi.: "he assumed that the original Septuagint was that which agreed most closely with the Hebrew text as he knew it . . . a step in the wrong direction."  A combination of the asterisk and obelus; see below, p. 71.  E.g. at Exodus 6:16, Gerson was substituted by Origen for Gedson. Whether his practice in this respect was uniform has not been definitely ascertained.  Hieron. Praef. ad Chron.: "quod maioris audaciae est, in editione LXX. Theodotionis editionem miscuit, asteriscis designans quae minus ante fuerant, et virgulis quae ex superfluo videbantur apposita." The Book of Job offered the largest field for interpolation: a scholion in cod. 161 says, Iob stichoi, ach' choris asteriskon, meta de ton asteriskon ,bs'.  See a complete list of these in Gardthausen, Griech. Paläographie, p. 288 f.  On an exceptional case in which he obelised words which stood in the Hebrew text, see Cornill, Ezechiel, p. 386 (on xxxii. 17).  A somewhat different view of Origen's practice is suggested by H. Lietzmann (Gött. gel. Anz. 1902, 5) and G. Mercati (Atti d. R. Acc. d. Sci. di Torino, 10 Apr. 1896: vol. 31, p. 656 ff.  This sometimes becomes a hook .  Prolegg. p. lix. sq.  Lietzmann proposes to read: Euagriou scholia eisin, hosa . . . arithmon, Or. de, hosa Origenen k.t.l.  The vertical bars denote, of course, the length of the lines of Cod. G. The lines of the LXX. column of the Hexapla, if we may judge by the specimen (p. 62 f.), varied in length according to the sense.  See the confused and inexact statement of Epiphanius, de mens. et pond. 18.  See Birt, das antike Buchwesen, pp. 100, 107 ff.  If the Hexapla was written in lines consisting of only one word like the Cairo palimpsest, this estimate is far too low; see Nestle in Hastings, D. B iv. p. 443.  See also the note at the end of the Scholia on Proverbs printed in the Notitia l. c.: metelephthesan aph' hon heuromen hexaplon, kai palin autocheiri Pamphilos kai Eusebios diorthosanto.  = Hpaul, Gregory, p. 449, Scrivener-Miller, i. p. 183 f.  See G. A. Smith, Hist. Geogr. of Palestine, p. 143 f.  Hieron. praef. in Jos.: "et sumptu et labore maximo indigent."  Ep. ad Sunn. et Fret. 2.  Adv. Rufin. ii. 27.  Identified by some with an Antoninus martyred three months before Pamphilus (Lake).  On antiballein and diorthousthai, see Scrivener-Miller, i.[p. 55.  Jerome says indeed (ep. ad Aug. ii.): "quod si feceris (i.e. if you refuse Origen's recension) omnino ecclesiae bibliothecas damnare cogeris; vix enim onus vel alter inveniatur liber qui ista non habeat." But he is drawing a hasty inference from experiences gathered in Palestine.  See c. v.  Jerome speaks elsewhere (in Esa. lviii. 11) of "exemplaria Alexandrina."  Fabricius-Harles, vii. p. 547 (cf. vi. p. 205).  This is however mere conjecture; see Harnack-Preuschen, i.:p. 442: "dass dieser Hesychius . . . identisch ist mit dem etwa gleichzeitigen Bibelkritiker gleichen Namens, ist nicht zu erweisen."  Das Buch des Propheten Ezechiel, p. 66 ff., the Hesychian group in Ezekiel is bs' klmphps, i.e. codd. 49, 68, 87, 90, 91, 228, 238 (Parsons). See also Ceriani in Rendiconti (Feb. 18, 1886).  For the Octateuch Mr McLean (J. Th. St. ii. 306) quotes as Hesychian or Egyptian MSS. H.-P. 44, 74, 76, 84, 106, 134, &c.  Cf. the scholion in cod. M at 3 Regn. iii. 46 enteuthen diaphoros echei ta anatolika biblia. The Lucianic text was also known as the ekklesiastike ekdosis (Oeconomus, iv. 548).  Oeconomus refuses to identify this person with the martyr and saint (iv. p. 498 n.).  Introduction to the N. T. in Greek, p. 138; c., the Oxford Debate on the Textual Criticism of the N. T., p. 29.  Introduction, p. 134 f.  Cf. F. C. Burkitt, Old Latin and Itala, p. 91, "Lucian's recession in fact corresponds in a way to the Antiochian text of the N. T. Both are texts composed out of ancient elements welded together and polished down."  Prolegg. p: lxxxiv. f.  See c. v.  Cf. his Prolegomena to Librorum V T. Canon. Pars prior graece (Gotting. 1883), p. xiv.  Or, as he denotes them, h, f, m, d, p.  A Hexaplaric reading due to Aquila; see Field ad loc.  Notes on the Heb. text of the Books of Samuel, p. li. f.  Mason, Persecution of Diocletian, p. 324.  Newman, Arians, p. 6 f.; Gwatkin, Studies of Arianism, p. 31 n.  Hort, Introd. p. 143.  On Lucian's work see the art. Lucianic Recension of the LXX. in Ch. Q. R. (Jan. 1901); E. Hautsch, Der Lukiantext des Oktateuch (in Mitteilungen des Septuaginta Unternehmens, Heft i., Berlin, 1910.  Praef. in Paralipp.  Cf. Hort, Introd. p. 142.
 Hieron. de virr. ill. 54.
 Cf. ep. ad Paulam.
 See D. C. B. art. Hebrew Learning (ii. p. 351 ff.).
 See D. C. B. art. Origenes, iv. p. 129 ff.
 Cf. Bp Westcott in D. C. B, iv. p. 99: "it was during this period (i.e. before A.D. 215) in all probability that he formed and partly executed his plan of a comparative view of the LXX. in connexion with the other Greek versions."
 Cf. Un palimpsesto Ambrosiano dei Salmi Esapli (Giov. Mercati) in Atti d. R. Accademia d. Scienze di Torino, 10 Apr. 1896; and E. Klostermann, die Mailänder Fragmente der Hexapla. The MS. does not supply the Hebrew column.
 In the MS. lanou appears in the third column, where it has displaced Aquila's rendering.
 MS. eurethes.
 Or Quinta? Cf. H. Lietzmann in G. G. A. 1902, v., p. 332: "die letzte Columne ist nicht, wie man anfangs glaubte, Theodotion, sondern die Quinta mit Interlinearvarianten."
 With marginal variants, eis to telos, psalmos (LXX.).
 With interlinear variant tois huiois (Th.).
 With marginal variants, eis to telos, psalmos (LXX.).
 MS. 1 a manu hemin (? Aq. Sym.).
 With interlinear variant heurethesetai hemin.
 With interlinear variant tais heurousais hemas (LXX.).
 MS. tais.
 With interlinear variant metatithesthai (LXX.).
 On selis, cf. Sir E. Maunde Thompson, Handbook of Greek and Latin Palaeography, p, 58.
 See also ib. 18 sq.; Hieron. Praef. in Paral., and in ep. ad Tit., c. iii.
 Used here loosely as = kommata, the kolon being properly a line consisting of a complete clause, and of 8--17 syllables: cf. E. M. Thompson, Gk and Lat. Palaeography, p. 81 f.; J. R. Harris, Stichometry, p. 23 f.
 In the earlier Cairo palimpsest even such words as 'l and me had each a line to itself; see Nestle in Hastings' D.B. iv. 443.
 Epikataskeuazein is insuper vel postea concinnare (Field, prolegg. p. xii.); cf. Dio Cass. l. 23 ta skaphe kateskeuase . . . kai ep' auta purgous epekateskeuase. Oeconomus (iv. 873), who regards the Tetrapla as the earlier work, understands Eusebius to mean only that Origen added to the LXX. the three columns containing A'S'Th' .
 Field, Hexapla, ii. ad loc.; cf. Hieron. in Psalmos (ed. Morin.), p. 66.
 It occurs (e.g.) in the Hexaplaric Syriac at 2 Kings 16:2.
 Cf. the practice of Aquila (Burkitt, Fragments of the Books of Kings acc. to Aquila, p. 14).
 Ep. ad Sunn. et Fret.
 See Driver, Samuel, p. xlvi.: "he assumed that the original Septuagint was that which agreed most closely with the Hebrew text as he knew it . . . a step in the wrong direction."
 A combination of the asterisk and obelus; see below, p. 71.
 E.g. at Exodus 6:16, Gerson was substituted by Origen for Gedson. Whether his practice in this respect was uniform has not been definitely ascertained.
 Hieron. Praef. ad Chron.: "quod maioris audaciae est, in editione LXX. Theodotionis editionem miscuit, asteriscis designans quae minus ante fuerant, et virgulis quae ex superfluo videbantur apposita." The Book of Job offered the largest field for interpolation: a scholion in cod. 161 says, Iob stichoi, ach' choris asteriskon, meta de ton asteriskon ,bs'.
 See a complete list of these in Gardthausen, Griech. Paläographie, p. 288 f.
 On an exceptional case in which he obelised words which stood in the Hebrew text, see Cornill, Ezechiel, p. 386 (on xxxii. 17).
 A somewhat different view of Origen's practice is suggested by H. Lietzmann (Gött. gel. Anz. 1902, 5) and G. Mercati (Atti d. R. Acc. d. Sci. di Torino, 10 Apr. 1896: vol. 31, p. 656 ff.
 This sometimes becomes a hook .
 Prolegg. p. lix. sq.
 Lietzmann proposes to read: Euagriou scholia eisin, hosa . . . arithmon, Or. de, hosa Origenen k.t.l.
 The vertical bars denote, of course, the length of the lines of Cod. G. The lines of the LXX. column of the Hexapla, if we may judge by the specimen (p. 62 f.), varied in length according to the sense.
 See the confused and inexact statement of Epiphanius, de mens. et pond. 18.
 See Birt, das antike Buchwesen, pp. 100, 107 ff.
 If the Hexapla was written in lines consisting of only one word like the Cairo palimpsest, this estimate is far too low; see Nestle in Hastings, D. B iv. p. 443.
 See also the note at the end of the Scholia on Proverbs printed in the Notitia l. c.: metelephthesan aph' hon heuromen hexaplon, kai palin autocheiri Pamphilos kai Eusebios diorthosanto.
 = Hpaul, Gregory, p. 449, Scrivener-Miller, i. p. 183 f.
 See G. A. Smith, Hist. Geogr. of Palestine, p. 143 f.
 Hieron. praef. in Jos.: "et sumptu et labore maximo indigent."
 Ep. ad Sunn. et Fret. 2.
 Adv. Rufin. ii. 27.
 Identified by some with an Antoninus martyred three months before Pamphilus (Lake).
 On antiballein and diorthousthai, see Scrivener-Miller, i.[p. 55.
 Jerome says indeed (ep. ad Aug. ii.): "quod si feceris (i.e. if you refuse Origen's recension) omnino ecclesiae bibliothecas damnare cogeris; vix enim onus vel alter inveniatur liber qui ista non habeat." But he is drawing a hasty inference from experiences gathered in Palestine.
 See c. v.
 Jerome speaks elsewhere (in Esa. lviii. 11) of "exemplaria Alexandrina."
 Fabricius-Harles, vii. p. 547 (cf. vi. p. 205).
 This is however mere conjecture; see Harnack-Preuschen, i.:p. 442: "dass dieser Hesychius . . . identisch ist mit dem etwa gleichzeitigen Bibelkritiker gleichen Namens, ist nicht zu erweisen."
 Das Buch des Propheten Ezechiel, p. 66 ff., the Hesychian group in Ezekiel is bs' klmphps, i.e. codd. 49, 68, 87, 90, 91, 228, 238 (Parsons). See also Ceriani in Rendiconti (Feb. 18, 1886).
 For the Octateuch Mr McLean (J. Th. St. ii. 306) quotes as Hesychian or Egyptian MSS. H.-P. 44, 74, 76, 84, 106, 134, &c.
 Cf. the scholion in cod. M at 3 Regn. iii. 46 enteuthen diaphoros echei ta anatolika biblia. The Lucianic text was also known as the ekklesiastike ekdosis (Oeconomus, iv. 548).
 Oeconomus refuses to identify this person with the martyr and saint (iv. p. 498 n.).
 Introduction to the N. T. in Greek, p. 138; c., the Oxford Debate on the Textual Criticism of the N. T., p. 29.
 Introduction, p. 134 f.
 Cf. F. C. Burkitt, Old Latin and Itala, p. 91, "Lucian's recession in fact corresponds in a way to the Antiochian text of the N. T. Both are texts composed out of ancient elements welded together and polished down."
 Prolegg. p: lxxxiv. f.
 See c. v.
 Cf. his Prolegomena to Librorum V T. Canon. Pars prior graece (Gotting. 1883), p. xiv.
 Or, as he denotes them, h, f, m, d, p.
 A Hexaplaric reading due to Aquila; see Field ad loc.
 Notes on the Heb. text of the Books of Samuel, p. li. f.
 Mason, Persecution of Diocletian, p. 324.
 Newman, Arians, p. 6 f.; Gwatkin, Studies of Arianism, p. 31 n.
 Hort, Introd. p. 143.
 On Lucian's work see the art. Lucianic Recension of the LXX. in Ch. Q. R. (Jan. 1901); E. Hautsch, Der Lukiantext des Oktateuch (in Mitteilungen des Septuaginta Unternehmens, Heft i., Berlin, 1910.
 Praef. in Paralipp.
 Cf. Hort, Introd. p. 142.