[Retros] Errata in "Chess Mysteries of Sherlock Holmes"?

[Larry Denenberg]

Hello.  I'm a complete novice at retro-analysis, having discovered the
field only by finding an old copy of Smullyan's book "The Chess
Mysteries of Sherlock Holmes".  I've been working my way through the
problems but I'm confused in a couple of places.  Otto and Joost assure
me that this mailing list is an appropriate place for beginners to ask
questions.  So here goes; I've put below three things that puzzle me.
Any help would be appreciated, either showing me the error of my ways or
confirming my suspicions.  If an errata list for this book (or for the
sequel) is available, I'd love to get a pointer to it.

Please reply to me directly, since I'm not (yet) a member of the mailing
list.  Thanks very much.

/Larry Denenberg
larry at denenberg.com
http://larry.denenberg.com/

----------------------------------------------------------------------

(All page numbers and quotations are from the 1979 edition of "The Chess
Mysteries of Sherlock Holmes".)

----------------------------------------------------------------------

"You Really Can't, You Know!"  (p. 38)

http://janko.at/Schach/Smullyan-SH/09.a.htm
r3k3/ppp3pp/6p1/P2Kp2P/1NB5/p5P1/PP3PP1/R7
White has just moved his knight.  Prove that Black can't castle.

Smullyan considers the possibilities for Black's last move and in each
case gives a reason that Black can't castle.  But when he comes to
Black's pawn at a3, he says only that if it moved last "it must have
been from a4."  I don't see why it couldn't have come from b4 (the White
knight wasn't there at the time!).  Smullyan's reasoning about this pawn
having made all its captures on white squares no longer applies.

I also think there's an easy fix:  Leave the diagram alone, but change
the problem statement to say that White has just moved his *bishop*.
Now if the pawn at a3 has just moved it must indeed have come from a4.

----------------------------------------------------------------------

In "Thoughts of a Logician" (p. 103)

http://janko.at/Schach/Smullyan-SH/26.a.htm
r3k3/P5P1/1P1P1P2/3PpK2/16/6B1/8
White can play and mate in two, but the mate cannot be exhibited.

Smullyan says "*either* Black can't castle *or* White can capture e.p.,
and there is no way of knowing which.  In either case there is a mate
in two, but a different one in each case."

But it looks like 1 Ke6 works in either case.  If Black does not castle,
then 2 g8(Q) mates.  If Black can and does castle, then 2 a8(Q) mates.

I think this is fixable by removing White's pawn at a7 from the diagram.

----------------------------------------------------------------------

"A Displaced Bishop" (p. 118)

http://janko.at/Schach/Smullyan-SH/31.a.htm
r3K2r/p1pq1ppp/1pnp1n2/b2P4/?5b1/?PNQ2N1/1PP2PPP/R3K2R
There is a White bishop on either a3 or a4.
Both sides can castle.  Determine where the bishop lies.

The problem is OK as stated, but I think it can be sharpened by removing
the hypotheses that Black can castle.  (E.g., replace the sentence "As
we stood watching, both sides successively castled" with "As we stood
watching, White castled.")

The fact that the Black king hasn't moved is used to prove that the pawn
originally on e2 could not have promoted.  But we already know that the
pawn originally on e7 must have reached e2 to promote on d1 or f1 with a
capture.  How did it get around the pawn from e2, which never left its
file?  It can't have been via two captures, since then the capture on d1
or f1 would make three captures, which too many.  Therefore the pawn
from e2 cannot have promoted since it must have been captured to make
way for the pawn from e7.  With this line of reasoning we never use the
fact that the Black king never moved and so can drop this hypothesis.