One of the most scholarly, authoritative, definitive and, indeed, expensive chess books of all time is published this month. Chess Problems: tasks and records, by Jeremy Morse (Faber & Faber, £30), has finally revealed to me the sort of thing great problemists are really trying to do. Take the first diagram, for example, composed by J C van Gool in 1977:


, , , ,

, , , B

b , z ,

H,h, , ,

X Z Nh,

B , V ,


It is White to play and mate in two. The solver's eye should be drawn to the bishop on a1, which would be pointless if it did not participate in a mate. So, unless the composer is being unfair, the first move must be with the rook on b2 or the king, to ensure the opening of the diagonal from a1 to e5.

The answer turns out to be 1.Kxc4! remarkably moving the king from total safety to expose him to 14 checks: Nb6, Qb5, Qxa4, Rxa4, Bb3, Nd2, Be2, Nxe3, Rc5, Nd6, Qc6, Qe6, Qf7 or Rg4. White's replies - you can work them out yourself - provide 13 different mates, only1...Qxa4 and Rxa4 having to share the reply 2.Rb4.

Those 14 checks and 13 mates are one of the many records detailed in this extraordinary collection of over 800 problems. Its completeness in detailing every record task, solved or unsolved, will make this book a problemists' bible.

If you want a two-mover in which a white pawn promotes to queen and knight on each of three different squares to give six mates, this is the book for you. Personally, I find it all rather over-technical, until we reach the humour of the last chapters on pawn promotion and length records. There is Blathy's mate in 257, and this one by Z. Maslar is a delight:

, , , v

, , , ,

, , , ,

, , , ,

, , , b

, , Zh,

hnhnhN n

zscd, ,

It's a helpmate in 8, which means that Black moves first and both players conspire to let White mate at move 8. The solution goes: 1.h1(R)! Kxf3 2.e1(N)+ Kg3 3.Bg4 f4 4.d1(B) f5 5.Bh6 f6 6.c1(N) f7 7.Qh7 f8(Q) 8.b1(B) Qf6 mate!