forgetting to include them in the text of their new copy, later wrote them in the margin.[5] In some cases, as we might expect, a different ink is used. The insular hand (= I), which we are assuming to be that of Johannes, corrects minor errors in these enlargements now and then.[6] This fact is entirely in accord with our hypothesis.

A number of enlargements omitted by the writers of the text were supplied not by them but by special correctors, who were assigned, it would seem, considerable portions of the manuscript to revise. Particularly important among these wide-ranging correctors are two hands that I will call r¹ and r². The former is a largish hand with some slight traces of Insular habits.[7] r² is very similar, and indeed may be merely a smaller variety of r¹. In the specimen that I have reproduced, as is true of both r¹ and r² elsewhere, correction by I may be observed.[8] In all, I detected, or thought I detected, five or six correcting hands, which sometimes supplement stretches of text written by others, sometimes supplement their own text, and, in all the cases under discussion, add notes of the author which were evidently in the margin of O. It is sometimes hard to be sure whether r is the text-hand or not. The point is not vitally important. The main fact is that several different kinds of correcting hand make, either in their own texts or in those of others, the kind of additions or enlargements with which we are specially concerned. However, as we have seen, we can still retain our hypothesis by supposing that I is the hand of Johannes, while r represents various correctors who copied from O enlargements added there by Johannes or at his direction.

But we have now to note an intimate connection between I and r. They collaborate on the same notes. Plate V (fol. 285^v) shows us an enlargement that begins in the hand (= r³) that writes the text. It extends through substantiam (1.3), then is succeeded by I (ex his—horum est), then returns (Ibi—superans), and finally gives way to I once more (dum—esse). The interesting possibility and enlargements taken from O. Possibly two or more stages are represented by O, r starting with an earlier, and I supplementing from a more complete form—but into that terra incognita of fresh hypothesis we need not enter. I's procedure, at any rate, seems exactly like that of r. Thus his practice of calling in a variety of r to complete a note too large for the space is paralleled by r⁴, the writer of the text on fol. 231^v (Plate II), who uses up a legitimate amount of his margin and then has r⁵ finish it, with signs of references, on the following page. The latter scribe uses a finer hand, and has no difficulty in completing the note with a decent margin to spare.