Like Susanna, and Bel and the Dragon, the Song of the Three Children formed, so far as we know, part of the original LXX text of Daniel, having a connection with it closer even than theirs. For while they take their places at the beginning or the end, this one is incorporated into the narrative of chapter iii. as one connected whole. Prof. Robertson Smith does indeed write (O.T. in Jewish Church, 1895, p.154), |these are perhaps later additions to the Greek version|; but this is only conjecture, and as such he puts it forward.
Until the correspondence of Origen with Africanus, the canonicity of these pieces does not seem to have been called in question by Christians who used Greek or Latin Bibles; nor do Greek-speaking Jews appear to have disputed the matter seriously. |Commonly quoted by Greek and Latin Fathers as parts of Daniel,| says Westcott (Smith's D.B., ed.2, i.713b). So Schürer (II. III.185), |Julius Africanus alone among the older Fathers disputes the canonicity of these fragments.| See also Bissell's admission on p.448 of his Apocrypha. But Jerome seriously called their canonicity in question (Præf. in Dan.), although he included them in his translation, with a notice that they were not found in the Hebrew. Polychronius, Theodore of Mopsuestia's brother, refused to comment on this piece because it was not part of the original Daniel, nor in the Syriac, ou keitai en tois Hebraikois e en tois Suriakois bibliois. In this latter respect it keeps company with the Catholic Epistles in the earliest stage of the Syriac N. T. (Carr, St. James, p. XLVII). But it gained a place in the Peshitto (D.C.B. arts. Polychronius & Polycarpus Chorepisc.). Buhl (Kanon und Text des A.T.,1891, p.52) says that the Nestorians recognise |die apokryphischen Zusätze zum Daniel als kanonisch;| and the Malabar Christians regard this, with its two companions, |as part and parcel of the book of Daniel.| (Letter to the writer from F. Givargese, Principal of Mar Dionysius' Seminary, Kottayam, 1902.) They formed part of the Sahidic, and probably other Egyptian versions of Daniel, which may be as early as century II.; as also of the Ethiopic and, seemingly, of the Old Latin (Swete, Introd.96, 107, 110).
It seems very difficult to prove that the Alexandrian Jews who used the LXX did not regard this piece as canonically valid; though how they reconciled their canon with the Palestinian one is not clear. Their frequent communication with Palestinian Jews must have brought any considerable discrepancy to the notice of both sides. F. C. Movers (Loci quidam Hist. can. V.T., Breslau, 1842, pp.20, 22) solves the difficulty by imagining that this and the other Apocrypha were similarly regarded both in Palestine and Alexandria, |vix credibile est alios libros a Palestinensibus inter profanos repositos ab Alexandrinis codici sacro adscitos esse.| Acts ii.10 proves the presence of Egyptian Jews at Jerusalem for Pentecost, and vi.9 that they had a synagogue there. This close connection must have brought their religious practices to one another's knowledge, and any differences, considered seriously important, could hardly have failed to raise disputes. Now Bleek (Introd. to O. T., II.303, Engl. transl., Lond.1869), says |the additions to Esther and Daniel were certainly looked upon by the Hellenistic Jews in just the same light as the portions of the books which are in the Hebrew.| And this seems to have been done almost without question, difficulty, or protest, although Alexandrian ideas must have been brought under the notice of the religious authorities in Jerusalem. (Cf. Meyer's note on Acts vi.9, and Jos. cont. Ap. I.7, as to regular intercourse between Palestinian and Alexandrian Jews.)
Professor, now Bishop, Ryle (Can. of Script. p.157) thinks that the amplification of Daniel, as of Esther, may have been tolerated because Daniel was not then deemed canonical. But we must remember that additional sections, though smaller in extent, appear in other books of the LXX, of whose canonicity there appears to have been no question, e.g. Job xlii.17, Prov. xxiv.22, I. Kings xvi.28, this last being taken from chap. xxii., though still left there. It has also been suggested by Prof. Swete (Introd. p.217) that the ktvvym were probably attached to the canon by a looser bond at Alexandria than in Palestine. However this may be, certain it is that this addition was frequently quoted or referred to by early Christian writers as if part of Dan. iii., without qualification or sign of misgiving, as may be seen in the quotations given in the chapter on Early Christian Literature,' p.76 sqq. Loisy's contention is a noticeable one (A. T. p.236), |Presque tous les auteurs catholiques, anciens et modernes, qui ont emis des reserves touchant l'autorité des deutero-canoniques, ont regardés ces livres comme inspirés. Ils ne les croyaient pas bons pour établir le dogme; mais cela est parfaitement compatible avec l'inspiration, attendu qu'un livre peut-être inspiré sans être dogmatique, et que s'il n'est pas dogmatique par son contenu il ne saurait regler le dogme.| But this contention savours somewhat of clever special pleading in order to evade the force of opposing evidence. Loisy, however, for a Roman Catholic, is a wonderfully frank and fair writer on these matters.
The explanation of the early mixture of non-canonical books with canonical, by reason of their having been kept as separate papyrus rolls in the same chest (Swete's Introd. p.225), seems not an unlikely one in the case of independent works such as Judith or Wisdom. But it appears to lose its force in the case of additions such as these, or those to the book of Esther. For the Song of the Three, Susanna, and Bel and the Dragon are hardly likely to have had separate rolls assigned to them; least of all this first piece, which fits into the middle of the accepted narrative, and is scarcely intelligible without it. Something more therefore is wanting to explain the inclusion of those portions in the Greek Bible.
Bengel's explanation (Gnomon on Matt. xxiv.15), that the apocryphal books in Latin Bibles were mixed with the canonical |pro argumenti affinitate,| though distinguished at first by marks (afterwards omitted) in the index, however likely so far as it goes, fails to account for their admission on so slender a plea into Biblical MSS. at all.
If the additions are to be regarded with Streane (Age of the Macc. p.161) as |specimens of fiction,| this one, more strongly than the other two, shews the pre-existence of the canonical Daniel; but it is very hard to understand how fiction' of this kind could be introduced into the Bible with no general protest, and ultimately come to be treated as of Divine authority; and this position is defended, even in these critical days, by the greater number of Christians in the world.
When the Council of Trent made the canon of Scripture co-extensive with the Vulgate, this, with the other additions, was of course included in the decree. But in the Roman Church up to the present day attempts have not been wanting to minimize the force of this decision, which, if it removes some difficulties, certainly introduces others. Outside the Roman Church the position of these book, in common with the rest of the Apocrypha, remains, as always, more or less insecure.
A. Scholz, in condemning the principle that Christians are tied to the O. T. canon, rather amusingly supposes: |Wenn Jemand sich bei den Juden jetzt als Prophet geltend machen und ein Buch schreiben würdem so müsste es nach diesem Grundsatz von den Protestanten als kanonisch wohl anerkannt werden| (Esther und Susanna, Würzburg, 1892, p.140). But such argument is mere polemic, which cannot be seriously taken into account in establishing the position of this or the other additions. Something is needed much deeper and more convincing in character.