47 Highpoint,
                                                                                                                            North Hill,
                                                                                                                            Highgate,
                                                                                                                            LONDON,  N.6.

                                                                                                                            24 January 1952.

Dear Bennett, 
                      I've been abroad for three weeks on a skiing holiday, and 
there doesn't seem to be anything which warrants inflicting another
Work Note on you all. As well, the more obvious features of the
Pylos tablets seem to have rather exhausted themselves. I haven't
heard from you for some time, but I hope that you are pressing on
your own Minoan studies and getting with some fundamental stuff
on Knossos.

                      I have been sketching out the more revised state of the
"Syllabic Grid" from note 15, with the alterations to Vowel 3 based
on the ending -u which I discussed in note 16. I'd like to just
think aloud in recapping the criteria I've so far used in deciding
which vowel a particular sign contains, and would be interested to
know if it tallies at all with the lines you have been working on.

Criteria for vowel-equations in Linear B.

1.	In the most typical form of the inflexion pointed out by Kober,
	where the addition of -jo / -ja alters the value of the
	connecting vowel, the nominative ends in Vowel 2   (to po no mo)
	and the oblique cases end in Vowel 1 (ti si ni mi). 
2.	In the less typical form of the same inflexion, the nominative
	may end in any of Vowels 3,4,5, reverting to Vowel 1 in the
	oblique case  (we-pi, nu-si, ne-ni,ra2-ri).
3.	Certain other occurrences of the ending –jo/-ja follow
	Vowel 3 instead of Vowel 1. This applies to some of the names
	introducing the Pylos As-Ad tablets.  Eg: -te-ja, -ne-ja, 
	-pe-ja, -ru-ja
4.	The vowel preceding the endings -jo and -ja, when they are
	the nominatives of masculine and feminine personal names respectively,
	consists almost exclusively of either Vowel 1 or Vowel 3: 
	eg: -ti-ja, -ni-ja, -mi-ja, si-ja, ------- -me-ja, -ne-ja,-e-ja, -re-ja. 
	This leads one to suspect that this name-formation maybe analogous
            in origin to the inflexion discussed under (1)-(3). 
5.	Signs with Vowel 1 do not occur before the sign o 
	(it is possible that signs with vowel 5 do not occur before –a, except in consonants). 
6.	The sign preceding the ending -u consist almost exclusively
	of Vowel 3.  Eg:            -we-u, -te-u, -ne-u, -pe-u, -ke-u, -re-u, -qe-u, -me-u.
7.	There is no absolutely hard-and-fast distinction between masculine
	and feminine in the personal name-endings and in words which vary 
	for gender, but there is a tendency for Vowel 1  (-ti, -ra, -ni )
	and Vowel 5  ( -ja, -wa, -sa, ra2), to be feminine, and for
	Vowel 2 ( -u, -to, -jo, -ro),Vowel 3 ( -te, -re, -ta) and Vowel
	4 to be masculine ( -wo). 
8.	The genitive sing. ending in -jo is regular after Vowel 1 ( -ti-jo, -ra-jo, -i-jo)
	and Vowel 2 ( -to-jo, -jo-jo, -no-jo, -ro-jo).
9.	The genitive ending in - o is regular after Vowel 3 ( -ta-o, -ne-o, -e-o) 
	and Vowel 4 ( -zo-o, -wo-o).
10.	The genitive ending after Vowel 5 is more frequently in -o 
	( -ja-o, -ra2-o, -ka-o), but there are some instances of  -jo ( -ja-jo, na-o).
11.	It is not absolutely clear whether Vowel 4 and Vowel 5 are to be
	kept distinct, but the size of the syllabary seems to suggest
	5 vowels. wi - u - we - wo - wa  is so far the only series of 5 to
	be filled up, and I believe you have some objections to including 
	-wa. 

                                                                              Yours sincerely, 

                                                                                              Michael Ventris