4:3. proselqwn o peirazwn eipen autw. The TR reads proselqwn autw o peirazwn eipen "the tempter came to him and said," according to C L, most cursives, one Old Latin MS, and the Harklean Syriac version. We read proselqwn o peirazwn eipen autw "the tempter came and said to him" with Tisch. Treg. Alf. West. Nest. UBS., according to S B W, some cursives, some Old Latin MSS, the Vulgate, the Peshitta Syriac, and the Bohairic Coptic. Most of the Old Latin MSS and the Old Syriac have autw in both places.

4:4. en panti rhmati. Others (West. Nest. UBS) have epi panti rhmati "upon every word," according to B S L and other uncials. We read en "in, on" with Lach. Treg. Alf., according to C D and some cursives. It seems that en ought to be preferred here after the ep artw earlier in the sentence, because it is likely that an original en would have been supplanted by epi as scribes automatically repeated the preposition from the first clause, but it is hard to see how en would have supplanted an orignal epi. The variation makes little difference exegetically. Both prepositions have a variety of overlapping senses, especially in Hellenistic Greek.

4:5. esthsen. The TR has the present isthsin "sets" instead of the aorist esthsen "set," in accordance with L W and the majority of cursives. We read esthsen with Lach. Tisch. Treg. Alf. West. Nest. UBS., in accordance with S B C D and a few cursives. The regular use of the present tense in the narrative (as in the preceding verb, paralambanei "takes") would have led copyists to substitute the present for the aorist here.

4:9. eipen. The TR reads legei "saith" instead of eipen "said" here, in agreement with L W and the majority of cursives. We read eipen with Lach. Tisch. Treg. Alf. West. Nest. UBS., according to S B C D Z and some cursives. Again, scribes would be inclined to substitute the present tense, in keeping with the style.

4:10. upage. Some editors (Gries. Lach.^text Word. Hodg.) add opisw mou "behind me" after upage "hence," according to C² D L Z and the majority of cursives, two Old Latin MSS, the Old Syriac version, and some MSS of the Coptic. (Most of the Old Latin have vade retro "go back.") We omit it, according to S B C* W and some cursives, some Old Latin MSS, the Vulgate, the Peshitta Syriac, and most MSS of the Coptic versions. The phrase was perhaps inserted as a remembrance of 16:23, where Jesus uses it in speaking to Peter. In the TR of Luke 4:8 the phrase upage opisw mou Satana is an interpolation.

4:12. AkousaV de. The TR adds o IhsouV after AkousaV de ("when Jesus heard") in accordance with C² L W and the majority of cursives, most Old Latin copies, the later Vulgate, and the Curetonian Syriac. We omit it with Tisch. Treg. Alf. Word. West. Nest. UBS., according to S B C* D and some cursives, the early Vulgate MSS and two Old Latin MSS, the Sinaitic Syriac and the Coptic versions. Probably the name was added here for the sake of clarity because this is the beginning of a Church lection.

4:17. metanoeite, hggiken gar. West.^marg omits metanoeite "Repent" and gar "for" after kai legein, according to the Old Syriac version and one Old Latin MS. The longer reading of all the Greek copies and other versions might be explained as an assimilation to 3:2 (the preaching of John), but the external testimony for this omission is so weak that it does not deserve serious consideration.

4:18. Peripatwn de. The TR inserts o IhsouV "Jesus" after Peripatwn de, with L E, some cursives, the later Vulgate, and some Old Latin MSS. We omit it with Gries. Lach. Tisch. Treg. Alf. Word. West. Nest. UBS. Hodg., according to S B C D and the majority of cursives, some Old Latin MSS, the early Vulgate MSS, the Peshitta Syriac, and the Bohairic Coptic version. The Curetonian Syriac also testifies to the absence of o IhsouV here because it inserts "our Lord" instead of "Jesus." As in verse 12, the name has been added for clarity at the beginning of a section.

4:23. en olh th Galilaia. The TR has olhn thn Galilaian (accusative case, without "in") with S¹ D W, the majority of cursives, and the Latin versions. We read en olh th Galilaia ("in" with the dative) with Lach. Tisch. Treg. Alf. West. Nest. UBS., according to S* B C and the Coptic and Syriac versions. The accusative construction was substituted as being more usual and idiomatic after perihgen.

4:23. perihgen en olh th Galilaia, didaskwn. Some add o IhsouV "Jesus" either after perihgen (Lach. and Treg.^text, according to S* C* D, Latin versions, the Sinaitic and Peshitta Syriac versions, and the Bohairic Coptic version) or before didaskwn (TR and Hodg., according to W, later uncials and the majority of cursives). We omit it with Tisch. Treg.^marg Alf. West. Nest. UBS., according to B, one Old Latin MS, the Curetonian Syriac, and the Sahidic Coptic version. The various positions of the word in the MSS which have it indicate that it is an insertion, and here again it seems that the name was supplied for the sake of clarity at the beginning of a lection.

4:24. daimonizomenouV. TR and some editors (Gries. Tisch. Treg.^marg UBS.^text) add kai "and, even" before daimonizomenouV "demon-possessed," according to S C² D W, the majority of cursives, the Latin versions, and the Sahidic Coptic version. We omit it with Lach. Treg.^text Alf. West. Nest. UBS.^marg, according to B C*, some cursives, and the Bohairic Coptic version. The kai may have been added by scribes who did not perceive that "those possessed with demons, and epileptic, and palsied" are the same ones already described in more general terms, "those oppressed with various diseases and torments" (basanoiV means "torments," not merely "pains" as in the RSV). If the kai is accepted into the text it should be understood not as a conjunction but as an epexegetical kai.