He was head of the Psychology Dept and an early user of computers. He was the moderator of our chess team.

He was very much beloved by every student that I talked too.

Da WHOLE SENTENCE about LACUNAE is oviusly used to make da chesgams comunaty AWARE dat u know what da word LACUNA means!

And <andrewjsacks> was too modest. He was sitting next to Donn and "only" drew his game against Fischer: Fischer vs A Sacks, 1964. And both <andrewjsacks> and IM John Donaldson saw that 39...Bxh4 instead of 39...Qe2 would probably have won. But, as <andrewjsacks> said in his notes to the game, "Yes, 39...Bxh4 was the move, all right--but as I think I said below, I was so pleased at that point as a 15-year-old kid to get a draw with the great Fischer, I did not notice it and simply went for the sure draw. C'est la vie."

The game appears to have been first published in the <Brooklyn Eagle>, February 29, 1940, where it is describe as "Marshall's latest". At Black's 10th move

click for larger view

Rogosin gave up a piece by taking the pawn on f4. The Eagle gives the move as <10...N(K3)xP>. This translates to 10...Nexf4, not 10...Ngxf4 as we have here. This move also appears in the <American Chess Bulletin> (March / April 1940, p. 45), and there's also the <Washington Sunday Star> (March 24, 1940), which probably just copied the Eagle.

Our <10...Ngxf4> is seen in <Chess Review> (August / September 1940, p. 128). It actually appears as <10...Kt(kt3)xP>.

So is this a typo, or did <Chess Review> have better information? I think there's an argument for the typo theory. The move <10...K(Kt3)xP, as given in CR, is ambiguous since that night could also take the e-pawn. The move should adctually appear as <10...K(Kt3)xBP>, as it does in Chernev's <1000 Best Chort Games of Chess>, p. 125.

What I'm thinking is that CR meant to copy the score from the earlier sources, but mistakenly changed "K3" to "Kt3". Because they didn't notice the typo, they didn't notice the ambiguity and thus didn't correct it.

Why has the CR version survived? In working with ACB and CR, I've noticed that this is quite common. My speculation is that it's not due to better quality or information, but to the simple fact that copies of <Chess Review> are more common than copies of <American Chess Bulletin>. Later writers were more likely to have CR, so used its versions. Database creators were more likely to have CR, so its version survived more often in the databases. And modern historical works, which are heavily dependent on databases, wind up using the CR version as well.