42.e4 was a sealed move and the players starting to analysing the game (together).

Botvinnik remarked: "A repetition of the past - I lost a similar ending thirty years ago in my match with Levenfish. Now I have mastered the way to play such endings."

Botvinnik: "Let us suppose that white has played 42.e4. Then there comes

42...a5 43.a4 e5! 44.Kf5 Kxh5 45.Kxe5 g4 46.Kf4 Kh4 47.e5 g3 48.e6 g2 49.e7 g1=Q 50.e8=Q Qf2+ 51.Ke5 Qe1+.

That is the way Levenfish won my queen."

Taimanov then opened the sealed enevelop, and wrote "and resigns" on his score sheet.

The king and pawns ending is the most nice I ever seen in a played game.

Botvinnik explained that the decisive white error was the transition of the K+R+P ending in the K+P ending:

31.Rxc1! (not 31.Rd4) Rxc1+ 32.Ke2 Tc2+ 33.Ke3 Rxa2 34.Rd4! and white should draw the ending.

Also 33.Rxe4 Rxa2 34.g4! followed by Kf1-g2 is better (according to Botvinnik).

it is included in Dvoretsky's Endgame Manual book exercise 1-7.

<Botvinnik remarked: "A repetition of the past - I lost a similar ending thirty years ago in my match with Levenfish. Now I have mastered the way to play such endings."

Botvinnik: "Let us suppose that white has played 42.e4. Then there comes

42...a5 43.a4 e5! 44.Kf5 Kxh5 45.Kxe5 g4 46.Kf4 Kh4 47.e5 g3 48.e6 g2 49.e7 g1=Q 50.e8=Q Qf2+ 51.Ke5 Qe1+.

Levenfish vs Botvinnik, 1937

But Botvinnik is thinking of a variation that *might* have happened after 58...Rg1 (instead of ...Rf2) 59.Re4 Kf6 60.b4 Kf5 61.Rd4 Rd1+ 62.Kc4 Rxd4+ 63.Kxd4 Kxf4 64.b5 etc

Unfortunately the parallel with the Taimanov game fails on two counts: with the black king on f4 it is close enough to g1 for the skewer trick to fail; and more importantly in this case White wins anyway as he will queen on b8 with check.