Winawer lost his game to Steinitz by time. We are informed that it was an exceedingly narrow shave, so much so, that the point of a penknife had to be employed to see if the finger of Winawer's clock had really gone past the hour on the thirtieth move. This kind of thing is decidedly unsatisfactory. It appears that Steinitz had at this time the superior game, but that is a fact altogether beside the mark, as advantages often end only in a draw, a ad also frequently go over to the other side, to say which is indeed mere commonplace. Several other games have been lost by time in this tourney, we mean directly lo6t, while quite a number of games have been lost through players finding themselves at critical stages with insufficient time. Thus considerations of merit and skill are altogether eliminated, for these stages may arrive quite fortuitously. Another evil result of the time limit slavery is, that players having positions admitting of deep and splendid combinations, are obliged, as often as not, to eschew them, for want of time to calculate and unravel. All this is very melancholy. The very base of such a system must be rotten. If players knew that at the time of their clook or glass reaching its limit, five or ten minutes grace would be allowed them, a great part of the evil would be eradicated, and that, whether or not the time of grace were debited to them.>

<OL v3 N33 (Sep 1883) G-182 p276/293>

on p277:

<Mit diesem knapp vor der Mittagspause (2 Uhr) ausgeführten Opferzuge überschritt Herr Winawer das vorgeschriebene Zeitmass von 2 X 15 = 30 Zügen für zwei Stunden, und wurde die Partie daher gesetzmässig Herrn Steinitz als gewonnen zuerkannt.

Nach der Hand kam es zwischen beiden Meistern zu einem kleinen Disput bezüglich der Correctheit des Springeropfers. Steinitz machte sich anheischig, „zehn gegen eins" zu setzen, dass er auch ohne die Zeitüberschreitung des Gegners gewonnen hätte. Winawer ging auf die Wette ein und das Probespiel nahm folgenden Verlauf:

29...Nxg3+ 30.hxg3 Bxg2+ 31.Rxg2 Rh6+ 32.Kg1 exf4 33.Nxc7 ( 33.gxf4 )
33...Qa7+ 34.Ne3 fxe3
( 34...Ng5 )
35.Nb5
( 35.Re2 )
35...Qe7
( 35...Qa8 )
36.Rf1 Qe6 37.g4 Re8 38.Nd4 Ng5 39.Rf5 Qe4 40.Qc4+
( 40.Qxe4 Nxe4 41.Rxa5 )
40...Kh8 41.Rxg5 e2 42.Nxe2 Qe3+ 43.Rf2 Qxg5 44.Bc1

und Weiss gewann nach einem längeren Endspiel, das nicht mehr aufgezeichnet wurde, auch die Wette. Sachlich ist dieser Ausgang übrigens nicht entscheidend, wenigstens hielt Herr Winawer seine Behauptung, dass er Remis machen konnte, auch nachher noch aufrecht.>

The subvariations added by me when engine spotted a significantly better move. Perhaps the two played a little loose knowing the result already decided?

Still, interesting enough to have on record.