Karpova: This correspondence game was played between the <Glasgow Chess Club> and the <Liverpool Chess Club> from October 1905 to October 1906.It was part of a small match, won by the Glasgow CC with 1.5-0.5.
Annotations (condensed) by <W. Gibson> of Glasgow, and <L. Hoffer> of London.
<LH> criticizes 7...Bb6, Lasker's Defense, as preventing complications and allowing White nothing other than to play for a draw. It would be even better to decline the Gambit instead, so that more freedom of action and room for creativity was left for both parties. If the game is prevented from becoming shallow, as in this case, it is entirely thanks to White.
According to <LH>, Black could have played <10...Nxe4>, e. g. <11.Bxf7+ 11...Kxf7 12.Nxe4 Rf8>, but 11.Nxe4 d5 12.Bg5 looks interesting.
<WG> thinks that 12...Be6 was probably the souce of all of Black's later evils. The Black ♘ is diverted, while the White ♗ becomes active.
<LH> suggests <15...c5>, but 15...c5
16.e5 dxe5 17.Nxe5 may be better for White than the game continuation.
Interestingly, the unnatural recapture 17.fxe3 is not annotated, but 17.Rxe3 (or Qxe3) looks more natural. It would also be easier to double ♖s on the e-file.
<LH> considers <19...b5> more forcing, but the Black position after the text move looks at least as fine.
<LH> calls 21.e5 <an ingenious combination>, as Black's counterplay on the queenswing came too late.
23...Kh8 looks a bit too passive. Someone (possibly Georg Marco) suggests the manoeuvre <23...Nc4> followed by <...Nd6> instead. But 23...b4 looks also interesting (24.d6 Qd8 25.Ne4 Bb5).
Not annotated, but interesting appears 34...Nef5 35.Rxf5 Qxf5 36.Rxd6.
Again not annotated, but Black may have better chances after 37...Ne7, planning ...Nef5.
Source: 'Wiener Schachzeitung', February 1909, pp. 61-62