Rev. 22:19 (KJV) And if any man shall take away from the words of the book of this prophecy, God shall take away his part out of the book of life, and out of the holy city, and from the things which are written in this book.
The Greek manuscript evidence for the reading "book of life" (biblou tes zoes) is hard to trace, for although it appears in the Textus Receptus, another reading, "tree of life" (tou xulou tes zoes), is in UBS-4 (without comment) and in the Majority Text in the NKJV Interlinear. Those who have commented on the verse, usually with obvious bias (cf. James White, The King James Only Controversy [Minneapolis: Bethany House, 1995], p. 66), state that Erasmus simply translated an erroneous reading from the Latin Vulgate, and that the error got into the KJV.
But regarding this Erasmian anecdote, Dr. Thomas Cassidy, pastor at First Baptist Church of Spring Valley (San Diego, CA), informs me in a private communication:
"Even so, to end the story here is to overlook the fact of the providence of God in allowing Erasmus to reach back to the Old Latin New Testament, via the Vulgate of Jerome, for his text does have some other witnesses - The Bohairic Version; Ambrose (4th Century); and the commentaries of Primasius and Haymo (6th and 9th centuries, respectively). So there is mss evidence, although slim, as the Peshitta and other early versions did not contain the Revelation at all. So we must conclude that God chose the Latin and a very few Greek mss to preserve the word 'book'.
"I know this is very slim evidence for the reading in the KJV, but as a very strong believer in the providence of God, I will stand by the evidence, no matter how slim it may seem."
Further details are given by Dr. Thomas Holland, who, in citing Dr. Jack Lewis' treatment of the subject, corrects Lewis' erroneous statement that the reading appears in no Gk. mss.. In fact, says Holland,
"It is found in the Greek manuscripts noted by H. C. Hoskier as 57 and 141. Nor is Lewis correct in assuming that there is no other textual evidence for the reading.
". . . [Libro (book) is the reading of the Latin mss.] Codex Fuldensis (sixth century); Codex Karolinus (ninth century); Codex Oxoniensis (twelfth to thirteenth century); Codex Ulmensis (ninth century); Codex Uallicellanus (ninth century); Codex Sarisburiensis (thirteenth century); and the corrector of Codex Parisinus (ninth century). It is also the reading of the Old Bohairic Coptic Version. Further, it is supported by Saint Ambrose (340-397 AD), by Bachiarius (late fourth century), and by Primasius in his commentary on Revelation (552 AD)" (Quoted from "Lesson 10: Textual Considerations," available at various internet sites that feature all of Dr. Holland's lessons in this series).
Another side point that may be of interest is that it is not quite that simple to say, "Erasmus took the reading from the Vulgate." The Clementine Vulgate does in fact read, "Et si quis diminuerit de verbis libri prophetiae huius, auferet Deus partem eius de libro vitae, et de civitate sancta, et de his quae scripta sunt in libro isto:", but in a modern critical text of the Vulgate like the Deutsche Bibelgesellschaft edition, one will find "ligno" (tree) rather than "libro" (book). Erasmus certainly could not have used the Clementine Vulgate--which became the standard Vulgate for over three centuries after it was first published in 1592--because he had been dead for over a half century when it appeared. This is not to say that the Vulgate that Erasmus knew did not (like the mss. Dr. Holland cites) read "libro" rather than "ligno"; but it does suggest that the easy dismissal of the reading as the fault of the Vulgate--when in fact the "standard" Vulgate bearing "libro" for "ligno" dates from several decades later, with Jerome himself apparently not responsible for the reading--is open to serious question.
While the external evidence for the KJV reading is not abundant, as Drs. Cassidy and Holland both concede, I find the internal evidence compelling.
Other than this passage, it is to be observed that "book of life" appears in Revelation six times: at 3:5, 13:8, 17:8, 20:12, 20:15, and 21:27. "Tree of life" appears outside this passage a total of three times: 2:7, 22:2, and 22:14. The "book of life" receives twice the mention of the "tree of life" for good reason: the "tree" appears to be a symbol of reward, while the "book" is symbolic of a person's very salvation. Notice the difference between 2:7 and 3:5 --the former pictures a believer who has already entered the holy city and is receiving fruit from the tree as a reward, while the latter pictures one just getting to enter the city (putting on "white raiment," which is the preparation for entering into His joys; cf. Matt. 22:11) and being assured of his entrance in the words, "I will not blot out his name out of the book of life, but I will confess his name before my Father, and before his angels."
That the meaning of the disputed words in 22:19 is "book of life" is confirmed by mention of exile from "the holy city." We are not in the realm of "reward" here (as we are in 22:12-14 , the last time the "tree" has been mentioned) but in that of soul-peril, as it is a serious thing to tamper with the words of the Lord. (Incidentally, the commentaries of Matthew Henry and Thomas Scott seem to confirm the soul-peril idea. Henry says that "he who takes anything away from [the word of God] cuts himself off from all the promises and privileges of it," while Scott notes that this is expressed "in the most awful manner" and that those who tamper with the canon "have abundant cause to tremble at this solemn warning.")
Let us consider how the readings may have arisen. "Book" could certainly be argued as a scribal miscopy from the line above in this same verse ("book of this prophecy"). However, "tree" could also be maintained as scribal in origin very early in the transmission, as the "tree of life" has been mentioned twice before in this chapter, while the "book of life" has not been since 21:27, so that a scribe might have expected "tree" here and so written it rather than "book." Although "tree" does weaken the force of the threat of damnation in the passage, it could be felt that, since the words "out of the holy city" also imply damnation, the "tree" reading allows a satisfactory sense in the context. Hence the "tree" reading could easily have been overlooked in early copies until it became the dominant reading, rather than what I take to be the genuine reading ("book") of the Textus Receptus and KJV.