In the example, da-/// should be allowed to come out of order after da-mo if it belongs. It doesn’t matter if there are two entries da-/// because they’re dependent anyway and it won’t throw any counting out. 5. I think the cf for the possible compounds should be made more dubious by a question-mark or two, because it shows a different level of speculation from the cf for the headless fragments. 6. Could K/N, P/Y and M/A, which look like presenting printing difficulties, not be written simply K, P and M? I agree that Kn, Py and Ma ought to be avoided because of their identity with the tablet references. Provided the Knossos entries have no alphabetic reference, or this is made sufficiently distinctive, I see no reason why we should bother to put K, P or M in the index at all. 7. I’m not sure about 4079 for 04-79 and 3132 for M132. While Scripta Minoa, admittedly with corrections, remain the main reference, I think we ought to stick to its exact numbering as printed there, though 04-79 might become 0479 (preferably not 04.79 as in your draft index), - provided this doesn’t conflict with the 479 in 4-figure reference as implied in your suggestions on classification. The result of these suggestions is shown in the slightly revised version of your “page 1” which I enclose. I agree that it would be an advantage to have a new classification of the Knossos tablets which would tie in with Pylos, and which one could use in the Index to give an immediate indication of the context. Miss Kober’s scheme is very fine as it goes, but it does seem rather too minutely worked out, and quite a lot of the X-series, for instance, can be attached to their proper series by taking a good look at them. Your suggestions are very simple, but I’m a bit alarmed at the prospect of yet another concordance, though it’s evidently necessary. It is obviously right to make the first alphabetical prefix agree with Pylos, and it’s a very good idea to apply these to the existing numbering of the tablets: it makes the Index very much more revealing. But it seems to me to be confusing if one ends up with two-letter prefixes which are similar to those of Pylos and identical with Miss Kober’s ones. I should prefer, myself, to see secondary classification within each series done in such a way that this identity did not result, and so that the Knossos and Pylos references were sufficiently different at first glance to avoid the need for the prefix K/N or P/Y in the Index at all. I should also prefer to see the Knossos numbers kept and not substituted by a 4-figure reference: 0004 for No 4 is, I think, a little complicated and wasteful of space, and I don’t think the fact of four digits as against the Pylos two is necessarily a clear enough distinction on the printed page. I’d suggest 3 possible alternatives: 1) To use one a one-letter prefix, at any rate in the Index (as shown). 2) To use numbers as the secondary classification: N5-53.3, M3-248.9 etc. 3) To use 2-letter prefixes as you suggest, with the secondary classification using letters vacant at Pylos, but with some distinctive hyphen or stroke to make a typographic distinction: Xb-12.3 Sp-487.1 Xb/12.3 Sp/487.1 But this still has the possible objection that the prefixes may clash with Miss Kober’s. However, I don’t feel very strongly about this, because obviously the classification which you propose in your letter would be perfectly workable and a great boon in whatever precise way it is recorded. I am only trying to foresee the objections that may be raised, for example, by Sir John, not that they need deter us. Yours, Michael Ventris PS: The reverse Index would we [sic] very useful indeed, & I agree that it should be compressed to exclude all references & indentations. Should it perhaps be annotated K P and M, though? * PS: I think it’s too much to expect that everyone will learn the signary order off by heart right away. I find it very helpful, in using the Index, to replace the page numbers by numbers corresponding to the places of the initials in the signary order: with a copy of the signary stuck on the wall, it makes reference much surer. I agree with your objections to a permanent reference numbering, but could something like this be evolved? It might also be an idea to have the signary diagram as a fold-out to the Index.