(1) Corrections by the original scribe, prima manu (p. m., 'first hand') as they are termed, can hardly be deemed variant readings. The penman, proceeding with his monotonous task rapidly and perhaps a little heedlessly, falls into some clerical error, which he immediately discovers and proceeds to set right ; in a few manuscripts (e. g. Codex D) by washing out the writing fluid, which was rather a kind of paint than ink, so that what he first copied can only just be perceived under his amended reading: in others, (e. g. Codex א), by placing points or some such marks over the letters or words he wishes to revoke (e. g. 'ÄÖ') .
(Here Scrivener has given the most obscure example possible at this point. The text here has faded so much that the mark is all but invisible now:)To give one instance out of thousands: In Acts. 4:3 (א), the text being την αυριον ην γαρ, while he was writing the first ην, the copyist allowed his eye to wander over αυριον to the second ην, and so he hegins to write γαρ : finding out his mistake as soon as γ is finished, he simply places a point over that letter to cancel it, and proceeds with αυριον ην γαρ as if no error had been made. ...
"(2) The next class of corrections is far more important. When a manuscript was completely written, it seems to have been subjected to two kinds of revision-process. (a) It was collated first with the master-copy from which it was derived, in order to eliminate whatever mere clerical blunders had not been noted at the time of writing; the person who executed this office was named the 'comparer' (ο αντιβαλλων), being usually the scribe himself.
(b) The second process was that of the διορθωτης, ("deorthotes") or corrector, seldom the same person as the comparer, whose business was to revise the text, often by the aid of a second manuscript varying a little from that first employed." (- F.H.A. Scrivener, p. xix-xxiii)