Game No. 165 in Steinitz's The Book of the Sixth American Chess Congress (1891), printed pages 213–215 (PDF pages 265–267); headed "Played in the second round, May 13th."

1. 9…Ng4 A novelty, but a somewhat risky one.
2. 10.Ne3 For 10 P—Q4, 10 PxP; 11 PxP, 11 B—Kt5 ch.; 12 QKt—Q2, would, we think, still more improve White's position.
3. 15.Qe2 Q— K sq. at once was clearly better,
4. 22.d4 By far better was 22 P—QKt3, 22 PxQP; 23 QxQP, etc.
5. 23.Be3 Black on the last move ought to have exchanged Pawns first, while White ought no more to have given him that opportunity.
6. 25…Nd3 Black's play since the opening moves is of a remarkable fine character.
7. 35…Qf7 35...Q—Kt sq. was superior, and the probable continuation was: 36 Kt—Kt3, 36 R—Q4; 37 KtxB, 37 QRxP; 38 Q—Kt4, 38 Q—B5; 39 QxQ, 39 PxQ; 40 K—B sq., 40 K—Kt sq.; 41 P—QKt4, 41 K—B2, and should win.
8. 38…Ra8 This turns out better than it deserved. The more simple plan was 38...R—Q4; 39 P—KKt4, 39 K—Kt sq.; 40 K—B2, 40 RxKt; 41 PxR, 41 K—B2 (not 41...RxP at once on account of the rejoinder RxP), remaining at least with a Pawn ahead.
9. 44.Rc6+ A grave error of judgment. He ought to have exchanged Rooks, followed by K—B3, and soon afterward by R—QR2, which would have won without difficulty.
10. 46…Kc3 The key to victory. Black's conduct of the ending is quite exemplary.
11. 56…Rg4+ Black might have won previously in a simple manner by R x KtP, but he most probably aimed at this exceedingly ingenious termination. See diagram.
12. 57.Kxg4 It was all over. If 57 K—B3, 57 R—Kt7 ch., followed by P—B7, wins equally.
13. 81…Qe5 A theoretical position and won according to theory.