In v.23 en to auto topo (O') are wanting as connecting words in B, but the reference to Bel in v.28 serves to consolidate the two portions of the story. A and Q also, as well as correctors of B, have an additional clause in v.24, which pre-supposes the former portion of the piece, a clause given in A. V. and R. V. The kai of me kai touton in O' answers the same purpose. Daniel's mocking tone at the end of v.27 agrees well with his sense of humour in v.7. Cyrus' ready compliance, too, in v.26 is only accounted for fully by the shock given to his idolatrous beliefs in the Bel part of the story. And so far the internal evidence argues for the unity of the piece. But it is noticeable that the Epistle for Tuesday after the Fifth Sunday in Lent in the Sarum and Roman Missals consists of the Dragon story only, beginning at v.29, with some slight introductory changes.
And Gaster's recovered Aramaic text (which he believes to have been the basis of Theodotion's Greek) consists of the Dragon story only. The notion that it had a separate currency is therefore, to a certain extent, supported; and this would still be the case, even if Gaster's text is not an original, but a translation.
If Gaster's Aramaic were really the basis of Th's version, it would follow that he did not confine himself to making a mere recension of the O' text, though he evidently availed himself of it as far as he thought proper. It is highly probable that this would apply to the Bel as well as to the Dragon story, although the corresponding Aramaic of the former is not at present forthcoming.
Neither the O' nor Th's original text seem to have been materially tampered with, either in the way of addition or omission. Each has some clauses not contained in the other: O' in vv.9, 15, 31, 39; Th in vv.1, 12, 13, 36, 40. Yet Westcott (Smith's D. B. I.397a, ed.2, 714a) thinks that some of Th's changes arose from a desire to give consistency to the facts. The change at the end of v.27, however, is hardly a happy one, kai eipen being put immediately after ho drakon, thus suggesting the idea that the latter drew attention to the fact that he was destroyed. The LXX. avoided this.
It is remarkable that Theodoret, in his Commentary on Daniel, comments on vv.1 and 2 of Bel and the Dragon (Th) only, treating them as the closing verse (14) of chap. xii., and introducing them with the words, houto plerosas ten apokalupsin epegagen ho prophetes; kai ho basileus Astuoges, k.t.l. This curious fact, combined with that of their omission from the O', points to some arrangement of the text with which we are not acquainted. Theodoret also refers to these same verses previously, in commenting on chaps. v.3 and x.1. Though he says nothing of the rest of Bel and the Dragon, he shews, by his referring in Ep. cxlv. (latter part) to Habakkuk's miraculous flight through the air, that he was well acquainted with the story, and approved of it.
The principal MSS. available are A, B, Q, G (vv.2-4 only), and D from v.21 to 41, which has recently reinforced our somewhat scanty uncial authorities.
The text of A appears to have slightly better Greek (vv.9, 10, 19, 21, 27); but the form machaires (occurs in Heb. xi.34 in A), if not a slip,  seems Ionic (Wordsworth's Greek Gram. § 16, Obs.), as has been already mentioned (Authorship,' p.193), and might perhaps be accounted for by Th's connection with Ephesus. The substitution of pros for to, however, in v.34 seems no improvement, A in this, as in several other instances (vv.10, 28, 35), agreeing with the O' reading. Taking, for convenience, B as the norm, we find that A's departures from it are somewhat larger than in the Song of the Three. In v.7 oude pepoken popote is added, as also in Q, to the description of Bel's inability to consume food. In v.11 daktulo is curiously substituted by A for daktulio; in v.13 katephthonoun for katephronoun. Both these are suggestive of carelessness or of error ex ore dictantis (Scrivener, N.T. Criticism, ed.2, p.10). In v.36 the substitution of cheiros for koruphes is peculiar. The alteration of gender in v.17, soai for sooi in its first occurrence, but not in its second, may come under the head of those "somewhat officious corrections" with which the editors of I. Macc. in the Camb. Bible for Schools (p.48) charge this MS., as likewise perhaps the reading paidion for teknon in v.10.
Q not unfrequently agrees with it in differing from B. It stands alone, however, in reading naon for hieron in v.22, and in omitting the last six words of v.41, perhaps as improbable when coming from Cyrus. Together with A, it contains an additional clause in v.24, putting words into Cyrus' mouth which connect the two stories together. G, having vv.2-4 only, contains no important variation. D (only from v.21 to v.41) contains in v.22 the curious word egdoma instead of ekdoton.
All things considered, the text of both versions may be said to be in as fair condition as in the canonical part of Daniel.
 There is clearly a slip in v. 35 of Daniel for Ambakouk, and probably in v. 11 of daktulo for daktulio, indicating some mistakes on the scribe's part, or errors in his copy.