Matt 7:4 απο

While most Greek manuscripts (including E G K L M S U V W X Δ Θ Π Φ Ω 047 0211 [Byz ca. 1490 mss] f35 565. 892. 1500. 2224) retain the preposition απο, which Greeven, Soden, Tischendorf (7th ed.), and Vogels approve, a few (ℵ B N Σ f1.13 33. 1424), followed by Bover, Lachmann, Merk, and Tischendorf (8th ed.), alter απο to εκ either by accommodation to the preposition of the preceding compound verb εκβαλω, by assimilation to the same expression in 7:5, or because the expression with εκ seemed more customary than with απο.
     Alford (1:70) and Soden (2:17) indicate that εκ arose from assimilation to 7:5, and Meyer (161) similarly notes, "With εκβαλω and ver. 5 before them, the copyists involuntarily wrote the εκ." Weiss (101–2) concurs, "But then all the newer [copies] resisted the mechanical assimilation of εκβαλω—εκ in Matt 7:4 (ℵ B), which anyway arose on account of the εκβ. εκ of v. 5 being so near, and similarly with the εξελθων εκ in 24:1 (B), where it is rather overlooked that the απο τ. ιερου does not at all belong to this verb, but rather to επορευετο, which is also why C X Maj. place it after the [verb]."
     Fritzsche (288) explains the origin of εκ but also the reason for απο thus: "The reading εκ has no power to stand randomly in the place of the common απο. And in fact εκ either came here from vs. 5 or απο was altered into the more common εκ by a scribe on account of its being less customary. But the difference is this, that εκ signifies that something has fastened to another thing out of which it is now moving, απο that something has been in a nearby thing from which it is now being separated." Also instructive is Cook's observation (64–5): "The word εκ is physically correct, but απο is better as referring to the intention. . . . In v. 5, εκ is generally adopted. The intention has been marked sufficiently by the old reading in the preceding clause; the act itself is now distinctly described."
     As compelling internal arguments corroborate the preponderance of external witnesses, there is little doubt that απο should be retained in the text.