C.G. Watney                                    R.I. Gunn                              
        The Observer, 1920                              Chess, 1941

    -------------------------------           -------------------------------
 8 |   |   |   |   |   |   |   |   |       8 |   |   |   | b | K |   |   | R |
    -------------------------------           -------------------------------
 7 |   |   |   |   |   |   |   |   |       7 |   | kt|   |   |   |   | P | R |
    -------------------------------           -------------------------------
 6 |   |   |   |   |   |   |   |   |       6 |   |   | KT| q |   |   | p | P |
    -------------------------------           -------------------------------
 5 |   |   | b |   | q | k |   |   |       5 |   |   |   |   |   |   |   |   |
    -------------------------------           -------------------------------
 4 |   |   |   |   | kt|   |   |   |       4 |   |   |   | KT|   |   |   |   |
    -------------------------------           -------------------------------
 3 |   |   | r |   |   |   | r |   |       3 |   |   |   |   |   |   |   |   |
    -------------------------------           -------------------------------
 2 |   |   |   | P | K | P |   |   |       2 |   |   |   |   |   |   | k |   |
    -------------------------------           -------------------------------
 1 |   | kt|   |   | B |   |   |   |       1 |   |   |   |   |   |   |   |   |
    -------------------------------           -------------------------------
     a   b   c   d   e   f   g   h             a   b   c   d   e   f   g   h

     White to play and mate in                    White to play and mate in 
           two moves                                    two moves

                     (Black pieces are shown in bold capitals
                          and white ones in small italics)

In the arcane world of chess problem solvers, or "problemists", a Meredith is a particular form of problem chiefly characterised by being describable as "white to play and mate in two moves", named after William Meredith, an American problem composer who popularised the form in the late 19th century.

The problems are given as diagrammed positions, invariably with a caption along the lines of "white to play and mate in two". From that position, play will be of the form:

1. <white move>, <black move>
2. <white move>++ (mate).

The first white move, called the key, is expected to be unique - the only white move on the board guaranteed to make checkmate possible after any subsequent black move and one more move from white. Any problem with two possible keys which both guarantee mate in two is called a cook, regarded as flawed, and quietly forgotten by composer and solvers alike.

The possible responses by black to the key are called the defences. For each defence there should be a unique subsequent white move available (a mate) which results in black being checkmated. In very complex problems, this requirement may be relaxed a little to allow a couple of possible mates after a particular black defence, but problems where each defence has a unique mating response will generally be regarded as superior. Each defence and its mate is called a variation, and the set of variations, together with the key, is called the solution of the problem.

A near-miss first move for white, which makes a mate possible after some defences, but not all, is known as a try. Problems with a high try to key ratio are generally favoured because this makes the correct key harder to find for the solver - a black defence which refutes each try must be found.

The mating possibilities in the starting position are referred to as the set play.

Often, the black king in the starting position has no possible legal moves. Any square which the king may move to after white's initial move is known as a flight square. Sometimes the correct key move will create a flight square for black, but this is only true for a minority of problems.

Generally speaking, for the best artistic effect (yes, composers and solvers will usually insist that these chess problems constitute an art form!) the key move should be as innocuous as possible. Sacrifices are by and large considered quite passée, taking a piece with white's first move is unusual (but still must be considered) and giving check is almost unheard of. If no black piece is taken by the key, no check given, and no threat made of immediate checkmate, it is called a waiting move. A waiting key which appears on first inspection to do nothing useful at all is generally thought to be the most satisfying kind. This puts black in zugzwang - chess lingo for a position where one side must move, but any move that can be made worsens that side's position - inevitably, in the case of Merediths, allowing white to pounce and give checkmate with the next move.

By convention, the starting positions given by the problems must be positions that could (however improbably) be arrived at by legal play in a normal game of chess. To avoid ambiguity in positions where a king and rook occupy the squares that they would at the start of a normal game, there is a convention, known as the castling convention, which states that if there is no internal evidence in the problem position that either piece has previously moved, and castling is legal on the board as shown, it is assumed that neither piece has moved, and castling may be part of the solution. This convention arose shortly after 1910, when the editor of Chess Amateur published a problem whose solution involved the threat of a castling move by White, provoking controversy and outrage amongst the readership, twenty five percent of correspondents declaring that the problem had no solution.

Given the contrived and often improbable nature of the problem positions, there is no direct relation between being good at solving Merediths and being good at chess. Where solving Merediths requires merely an exhaustive tactical analysis going only two moves deep at most, playing chess proper demands positional judgement and strategic skill, neither of which have any relevance at all to Merediths. Nonetheless, solving these and other types of problem may be a useful exercise, training the player to spot mates, examine all possible moves, and keep an eye out for the unexpected.

Noted problemists include Vladimir Lenin, Jacob Bronowski, Vladimir Nabakov and Philip Marlowe.

(The solutions to the two Merediths at the start of this writeup will appear here in a couple of days - promise!)

Terms, information, and problems from: Pick of the Best Chess Problems Compiled: B.P. Barnes, Elliot Right Way Books, 2nd edition, Reading, UK, 1991.