The probability of a Semitic original lying in the background of this piece, has always been considerable. Those who have maintained Greek as the original language, have generally spoken a little less confidently with regard to this than with regard to its two companion pieces. So Bissell writes (p.443), though a supporter of the Greek (p.43), "undoubtedly more can be said in favour of such a theory" [of a Semitic original] "than for a similar one in respect of the two remaining additions." And since M. Garter discovered in 1894 an Aramaic text, the grounds for deeming the Greek to be the original, though not set aside, have been partially undermined. Schürer, however, in Hauck's Encycl. (I.639), appears to think that this is translated from Th, and not vice versâ, as Gaster claims. In his third German ed. of H.J.P. (III.333) he agrees with Gaster in deeming tvkvs to be Th, but considers the Aramaic to be a rendering of Th's Greek, taken into the tenth-century Chronicle of Jerahmeel.

It must be confessed that the existence of two Greek versions increases the probability, though it does not prove the existence, of an original in another language. It does not seem likely that Th would have revised the O' of the additions in the same way as the canonical part, unless he had a similar basis to go upon in both cases. If not, why, and on what authority, did he alter the additions at all? And this consideration applies to the other two, even more than to the one we are dealing with, inasmuch as the version of Susanna and of Bel and the Dragon involved more numerous changes. Irenæus' statement that Theodotion "hermeneusen," taken strictly, would of course always imply an original to translate; but Irenæus may only have been thinking of the particular passage from Isaiah which he refers to (III. xxiii.).

Many phrases may be instanced which point to a Semitic original, or at least fit in well with the theory of its existence. Towards counterbalancing this there is a much smaller number which may be thought to tell in the opposite direction. But in the main, as Cornely truly writes (op. cit. p.420), "accedit hebraismorum frequentia quum in Alexandrini tum in Theodotionis versione." [20]

It is to be observed, however, that the names of the Three are Grecized from their original Hebrew nomenclature, [21] although their Babylonian names are employed in Dan. iii., and adopted by O' and Th in the canonical portions, both before and after the apocryphal episode. An apparent exception occurs in v.23 of O', where clauses of that verse and of v.22 have been transposed and slightly altered. Here Azarias occurs in the same form as in the apocryphal portion. But this isolated use of the Hebrew form of his name has probably been brought about by the insertion of our piece into the chapter, the same form and phrase, tois peri ton Azarian, being found in v.49 of both Greek texts. A like phrase occurs in Ezek. xxxviii.6, and in Acts xiii.13. The order of names, too, differs in this Addition from their order elsewhere, the two last changing places, thus bringing Azarias (Abed-nego) into the middle. It is remarkable that he is twice, vv.2 (25) and 8 (49), placed as if he were the leading member of the trio, in the former verse as uttering the prayer, in the latter as heading the party in the furnace; and so also, as pointed out above, in v.23 of O'. This last fact, however, is counterbalanced in the same version by all three being named in v.24 as praying, Azarias not there figuring as the sole speaker. These small indications certainly point to some ancient distinction between the uncanonical insertion, as we have it, and the body of the book.

E. Philippe (in Vigouroux' D. B. II. p.1266) argues for Hebrew and not Greek originals, because of the existence of two Greek versions, neither of which, he says, appears to be a revision of the other, containing hebraisms suggestive of a Hebrew original. But as regards the Song of the Three, this statement, that neither version is a revision of the other, must be regarded as more than doubtful. He also says that the Chisian and Syro-Hexaplar MSS. contain critical signs of Origen, revealing a Hebrew text, and in 87 (Chisianus) at xiii.1-5, A', S', Th' indicate Aquila, Symmachus, and Theodotion, all translators from the Hebrew. This last point, however, may not stand as to the Song of the Three (see note in Kautzsch, p.176) so far as Aquila is concerned. For Origen, in his letter to Africanus, seems to imply that Aquila's rendering did not contain the Song: Houto gar Akulas douleuon te Hebraike lexei ekdedoken -- § 2.

Jerome's words in the Vulgate, after v.23, "quae sequuntur in Hebraeis voluminibus non reperi," are very guarded, not absolutely denying the existence of a Hebrew text, but merely asserting that he has not met with it. Cod. Amiatinus, however, has 'non repperiuntur,' an expression which asserts more comprehensively the absence of this passage in his time.

The following are some specific indications of language which appear to be of sufficient interest to be noted separately:

v.27 O', Th. Dikaios ei epi pasin = tsryq l rendered by epi in Dan. ix.14 (in both versions) and in Neh. ix.33. Dikaios epi also occurs in Bar. ii.9, in that part of Baruch which is almost certainly a translation from the Hebrew. Ball (Speaker's Comm.) gives a similar phrase from the Iliad, and Bissell a still more apposite one from Il. iv.28, to shew that it is not unknown in pure Greek. Gaster's Aramaic has simply l not l

v.30 O', Th. Hupakouo governs the genitive correctly, but suntereo, coupled with it, is made to govern the same noun. Exigencies of translation might easily cause this awkwardness, but hardly original Greek composition.

v.31 O'. Kai nun = vth So translated in II. Chron. vi.16, 17 at the beginning of the verse, as here; it occurs again in vv.33 and 41 in both versions, as also in ix.15,17. It is not a very natural beginning of a Greek sentence.

v.32 O', Th. Why apostaton, a title which does not seem very applicable to the Babylonians? It may be merely a rendering of mrd as in Ezra iv.12, 15. The Vulgate here has 'prævaricator.' In Gaster's Aramaic the verse is different. But cf. use of apellotriomenoi in Eph. ii.12 of those who had never belonged to Israel.

v.33 O', Th. Ouk estin hemin anoixai looks very like a translation of 'yn lnv, an idiom used in II. Chron. xxxv.3, 15 in the sense of cannot,' followed by a verb in the infinitive. Cf. Heb. ix.5.

v.34 O', Th. Eis telos = lklh or lntsch as in II. Chron. xii.12, Ps. xv.11. Diaskedases sou ten diatheken. This curious expression may be the rendering of such a phrase as that in I. Kings xv.19, hphrh 't brytk, there translated by the same words.; also in Jer. xi.10.

v.36 O', Th. Astra tou ouranou, as in viii.10, xii.3, both O'.

v.37 O', Th. Tapeinoi en Did the translators read vkl for tkl?

v.38 O', Th. Karposai. Cf. Lev. ii.9, l1, qtr 'sh being similarly translated. Karpoo is also used in the same sense in I. Esd. iv.52. Deissmann has an interesting study' of this word in his Bible Studies (Eng. transl., Edinb.1901, p.135).

v.40 O', Th. Enopion . . . opisthen = 'chry . . . lphny. Ektelesai is thought by Ball to have arisen from some confusion between klyl and kll, but this is dubious. Marshall (Hastings' D. B. iv.755b) suggests sls in Kal or Piel.

v.44 O', Th. Endeiknumenoi, Grotius (in Critici Sacri) says "Expressit Hebræum hr'h quod est in Ps. lx.3 (5) et alibi." The verb is so translated in Exod. ix.16.

v.49 O', Th. The apparent Grecism of of hoi peri ton Azarian occurs in the LXX of Ezek. xxxviii.6 and elsewhere. Sunkatebe hama, Ball suggests yrd 'chry from Ps. xlix.18. Gaster gives nchyt m. Exetinaxe, Gaster characterises as a "senseless" rendering of v'ytstnyn "and it cooled down," which word certainly gives an excellent sense.

v.50 O', Th. The well known " crux" of pneuma drosou diasurizon appears in the Aramaic as dy tynsv' tl' krvt' which Gaster translates "as a wind that blows (and causes) the dew (to descend)."

v.51 O'. kai egeneto = vyhy

v.54 O'. Doxes tes basileias, cf. Dan. iv.36 (33) Th' , timen tes basileias. yqr mlkvt is the Aramaic in both places. thronou doxes, as in Jer. xiv.21. thronos is used of God's throne in Dan. vii.9, end.

v.59 O', Th. Ouranoi = smym (not in Gaster's Aramaic).

vv.64, 68 O'. Repetition of drosos, and vv.67, 69 O', of psuchos, suggests possible difficulty of a translator, causing him to fall back on same word.

vv.65, 86 O', Th. The different senses of pneumata point to rvchvt as the underlying original of both.

vv.87 O', Th. Tapeinoi te kardia Luther renders "elend and betrübt sind," since these words, if of literal and immediate application, would indicate the depression of the Babylonian exiles; and so would tell in favour of a Semitic original, Greek being unfamiliar to them.

vv.88 O', Th. Ek mesou kaiomenes phlogos, cf. Dan. iii.21, 29; vii.11 (yqd, Chald. in first and third of these cases, and also in Gaster's Aramaic of this piece).

vv.89 O', Th. Exeileto does not seem a very suitable word, as they had not yet been into hades. It may be a translation of ys as in Jer. xlii.11, if from a Hebrew original. syzvn' is given by Gaster as the original of both exeilato (Th) and errusato.

vv.90 O', Th. Hoi sebomenoi, used of proselytes of the gate in Acts xvii.17, may have this meaning here also, as coming last, and in connection with ton theon ton theon, a possible reference to the "gods of the nations." Gaster's Aramaic has nothing answering to sebomenoi. Grotius suggests "yr'y 'lhym" ut Job i.1, 8, ii.3," where theosebes is the word.

The writer deems the evidence of language to point on the whole to a Semitic rather than to a Greek base. The difficulty of balancing the indications however of the original language is shewn by the names of important authorities which may be ranged on either side, Ball, Rothstein, and Swete regarding the Semitic as probable; Westcott, Schürer, and Fritzsche holding a similar opinion as to the Greek.

When a Semitic original is pronounced for, the further question arises, was it Hebrew or Aramaic? The grounds unfortunately appear too indecisive to warrant a distinct choice between these alternatives.

STYLE.

This is the only one of the three Additions which takes a devotional and poetical form. The Song has perhaps exceeded the others in the great estimation accorded to it. The frequent liturgical use made of it is both a sign and a cause of this.

The style of the Greek is Hellenistic, and is not out of character with the versions of which it is a part; nor in particular with the Book of Daniel with which it is incorporated. It is spirited, interesting, and agreeable, mainly Hebraic in the character of its thought and cast of its language.

The Prayer may possibly be accused of the needless repetition of similar sentiments; especially in vv.4, 5, and 8 as to God's truth and justice; and in vv.6 and 7 as to Israel's disobedience, which are somewhat over-insisted upon. But perhaps this may be attributed to earnest pleading. It is instructive to compare and contrast Daniel's Prayer, chap. ix., remembering that a different person would naturally have a different style; a consideration which may also help to account for the change we are conscious of when we pass from the prayer of Azarias to the Song which purports to be the composition of the Three.

The principle on which pas is inserted in some verses and omitted in others does not seem clear. Rhythmical considerations do not sufficiently account for it. Something other than style seems to have influenced its use; but what that something may have been it is difficult to discern. Nor does the principle seem clearer in the Aramaic than in the Greek.

The poem has a simple yet majestic structure, with a refrain apt to linger in the ear, either in Greek or English, Eulegeite, humneite, kai huperupsoute auton eis tous aionas, "Bless ye the Lord, praise Him, and magnify Him for ever." In Gaster's Aramaic the refrain is slightly varied, llm' being used where God is addressed, vlm' where His creatures are exhorted. Dr. Gaster understands the former to mean "for ever," but the latter "in the world." [22] This distinction, if a just one, is entirely obliterated in the versions. In the Vulgate however the refrain sounds less agreeably, for "superexaltate" is a cumbrous word for frequent repetition. It is one of those exaggerated compounds of which the translator of Daniel seems to have been too fond, such as "superlaudabilis," "supergloriosus " (v.52), "deambulo" and "discoöperio" (Sus. vv.8, 32). This inconvenience was evidently felt in liturgical use, as in the Roman Breviary and Missal the repetition of "superexaltate" is avoided. Psalm cxxxvi. affords a biblical instance of a refrain similarly repeated at the end of each verse; and Deut. xxvii.15-26 may be regarded as containing a liturgical repetition of another species.

The use of a symbolic multiple of 7 in v.24 (47) accords well with a similar practice in Daniel iii.19, ix.24, and x.2, 13. The number 3 itself (v.28) may also be symbolic; but this is merely continued from the canonical part of the story, being quite of a piece with it. No other numbers occur.

There is a remarkable resemblance between the natural objects mentioned in Ecclus. xliii. and in the Song. Especially v.22 [23] of the former is like v.27 (50) of the latter in its leading idea. The furnace, kaminos, is also named in v.4 of the Ecclus. passage; and the aim of glorifying God is most prominent in both. But the resemblance in style to Psalm cxlviii. is not so great as has sometimes been imagined. (See what is said on this point under 'Authorship,' p.26.) On the whole, the style of the work, whether supplicatory, narrative, or poetic, is well suited to the purpose for which it is designed; and although the influence of previous writers is evident, the manner of the author is not that of a mere imitator of their compositions. He has a form of his own in which to present his subject.

[20] Dr. Julian (Dict. Hymnol. p. 134) has the following strange sentence as to Benedicite, "It is not in the Hebrew version (sic) of the Scriptures, and on this ground, among others, it is omitted from A. V."

[21] G. Jahn in loc. thinks this fact an indication of a later hand, as shewing that they severed themselves in the furnace from contact with heathenism, and were giving themselves to intercourse with Jahwe alone. But surely an interpolator must have been aware that this was their attitude from the outset.

[22] Proc. Soc. Bibl. Archaeol. 1895, p. 80.

[23] In the Hebrew of this verse the parallel is less striking.