47 Highpoint, North Hill, Highgate, LONDON N.6. 6 Sept 51. Dear Bennett, Here is the last of the Pylos Work Notes. The Figures 9 & 10 on consonant equations have still got to be done: I meant to get them done by now, but I couldnt quite make it. I'm going to Turkey next week to combine sightseeing with the Orientalist Congress, at which Bossert says Minoan will be discussed amongst other things, and the main bones of the "inflexion" Note is at least finished in time for that. I'd be glad any time now to compare Notes with the conclusions of your thesis: Myres is still very sceptical about inflexion (is the Knossos material much less helpful in this respect, or is he just obtuse?!). Thank you for your letter about the Statistical analysis. I quite agree with your compromise arrangement of the tables. (In parenthesis, I violently disagree with your preference for the order of diagraphs!! Every time I have wanted to use the diagraph table for any specific purpose it is the sequent pattern which I have needed. But on the assumption that with a little extra trouble the pattern can be extracted from either arrangement, I am prepared to let it go.) It is only in the case of Table I that I am not quite clear. I quite agree with the arrangement you give, but I can't understand your comments (i) that my "Max poss. total of all sign-groups" is clearly a "by occurrence" calculation; (ii) that the different methods of gaining the total on right & left correspond to maxima & minima. My "max possible total" was only intended to give a theoretical total if all the apparently dependent sign-groups were in fact discrete too: but it doesnt mean very much, admittedly. On your other way of looking at it, please elucidate, as I'm a bit fogged. On the question of my helping in the count, please be quite frank:- if I can genuinely help, I'd be glad to, though this summer's work has rather upset my home life, & I'd like to ration my Minoan, as you, during the next twelvemonth. But if it's really a tedious duplication of work for you to brief me & copy the index, & you only suggest a collaboration out of good nature, then think twice about it. But on the assumption that I do help, this is what I think I'll need:- (i) A sketch of the tables as in your letter (that is, if they still change at all) & a note of all the various 'rules'. Unfortunately I wrote my last letter to you over a rather sun-bemused week-end, and I didn't keep a copy. So some of my suggestions I am afraid I can't remember! (ii) An index of each set as your Pylos index, on any convenient size of paper: not cards, which I used before largely in order to rearrange the index to Myres order: now you've won on that one, it becomes academic. Myres is probably a bit nettled about your signary, but it seems sensible to adopt it since his alternative is a hodgepodge of various sorts of signs & can't be used for our purposes decently. (iii) In the Knossos index, of which i should very much prefer to see your version, perhaps the sign-groups whose readings correct those of Myres could be underlined.