It has been suggested by Prof. Rothstein (in Kautzsch I.174, 175) that the prayer of Azarias, the intermediate narrative, and the Song itself, were not all written at the same time. But this view is based purely on internal probability, and derives little or no support from any of the MSS. or versions, unless the introduction of titles in the Arabic after v.28 (51), and in some Greek copies to the prayer of Azarias, be thought to give it countenance; yet these may have crept in from their convenience for liturgical use, and so be accounted for merely on practical grounds.
To base this separation, however, on a supposed disagreement between v.15 (38) and vv.31 (53), 62 (84), is certainly insufficient cause, as Ball points out (307b), for assigning Prayer and Song to different writers (see Chronology,' p.67). But the observation that the narrative passage between the Prayer and the Song fits in well after the canonical v 23 seems a stronger basis for supposing that the prayer is a later introduction than the Song. Rothstein points out (p.181, note d) that v.1 (24) in Th has relation to the Song, but not to the Prayer, and originally, as he imagines, took the place of the present v.28 (91) of similar import. Corn. a Lap. notes of v.1 (24) |est hysterologia.| This view is also mentioned with favour in Charles' article on Apocrypha in the 1902 vols. of Encycl. Brit. (cf. For whom written,' p.36).
It is observable also that the statement of v.26 (48) is not a mere repetition of that in v.22, but refers to the scorching of the onlookers, while v.22 speaks of those who executed the king's order.
The repetition of the same invocation at the commencement of the Prayer and the Song is noteworthy; if the two are not contemporary, it has probably been borrowed by the composer of the Prayer. But the difficulty (often magnified) of reconciling the statements of v.15 (38) with the Jews' civil and ecclesiastical condition at the time of Daniel iii. wears quite a different aspect if the Prayer is regarded as an interpolation of later date by another hand. Altogether this theory of the interpolation of the Prayer is surrounded with a considerable air of probability.
Five extra verses are interspersed in the Syriac of the Song, calling upon the hosts of the Lord, ye that fear the Lord, cold and heat (the winter and summer of our Benedicite), the herbs of the field, and the creeping things of the earth (Churton's translation). Of these |frigus and aestus| is in the Vulgate, taken from Th. The source of the others is unapparent, though creeping things would very naturally follow beasts and cattle, as in Gen. vii.14.
The present ending of the Song, after the usual refrain in the middle of v.66 (88) is of a laboured nature with a decidedly |dragging| style. It certainly has the appearance of being an afterthought, added by some not very skilful composer, who fancied the original termination to be too abrupt, and thought he could attach an appropriate supplement. But of this theory no external evidence is at present forthcoming.
Th agrees with the O' text much more closely in this than in the other additions. Most verses are the same, word for word; and many others have but the slightest variations. He makes a few small omissions, as in (Greek) vv.24, 40, 67, 68; but in general he follows O' exactly. Even vv.67, 68, are contained in A, in both places, in Daniel and in the Odes at the end; also they are in the Turin Psalter, though omitted in the Veronese (Swete's LXX). As they are found, with a little difference in the O' text, they may have fallen out of B and Q accidentally. The identical refrain at the end of each verse would naturally facilitate an error of this kind.
The principal MSS. available for Th's text are the same as those for the canonical part of Daniel, A, B, and Q. G fails us here, as in other passages, except from vv.37-52, in which its variations are unimportant.
Taking B as the groundwork, A's changes are not generally of serious moment, excepting in the case of the two inserted verses, 67 and 68, and the transposition of vv.73 and 74. Otherwise they chiefly consist of small insertions or omissions which do not materially affect the sense (vv.36, 81); varying forms from the same root such as huperainetos for ainetos (v.54), eulegemenos for eulogetos (v.56). The correctors of B in v.38, though unsupported by the chief codices, certainly seem right in substituting oude for ou. Q's variations not unfrequently agree with A's; where they do not, they are scarcely more important, and often partake of a similar character. In v.88 a synonym is substituted, viz., esosen for erusato (2nd). In the few verses covered by G, B is generally agreed with; a change of case, autous instead of autois, appearing in v.50.