Edgehog Chess was invented in 1966 by John Driver. It was likely invented as a way to make use of the newly invented Edghog Piece. Actually, John Driver invented two versions of Edgehog Chess, with two different versions of the Edgehog piece. This article also describes a third version of Edgehog Chess invented by Peter Aronson.
This version of Edgehog Chess is very much like standard orthodox chess -- the only change is that each side's Queen is replaced by an Edgehog. The Edgehog moves like a Queen, but must either start or end its move on an edge square. Pawns promote to Edgehog, Rook, Bishop or Knight.
This version of Edgehog Chess is a bit more changed from International Chess, as all of the pieces except for the King and Pawns have been replaced by fairy pieces.
The Queens are replaced by Locusts, which jump on Queen lines over opposing pieces only, landing on the first square past the opposing piece (which must be empty), capturing it. Locusts may only move to capture.
The Rooks are replaced by limited Edgehogs, which move like Queens, but if they start their move on an edge square, must end up on a non-edge square, and if they start their move on a non-edge square, must end up on an edge square. Presumably, the King castles with the limited Edgehogs as if they were Rooks, even though that violates the limited Edgehogs movement restriction.
The Bishops are replaced by Reflecting Bishops, which can bounce off of the side of the board at a 90 degree angle and keep moving. A Reflecting Bishop can bounce off of multiple sides of the board in a single move, but may not end up on the square on which it began the move.
The Knights are replaced by Nightriders which can make repeated Knight moves in the same direction, as long as all but the last square landed on are empty.
A Pawn promotes to a Locust, Reflecting Bishop, Nightrider or limited Edgehog.
While many of the individual pieces in Edgehog Chess II are fairly powerful, they are all awkward in one way or another, making it hard to achieve mate. Edgehog Chess III makes a few changes to Edgehog Chess II that are designed to alleviate this problem.
First, the limited Edgehogs in the Rook positions are replaced by regular Edgehogs. They can still castle.
Second, instead of starting in the Queen's position, the Locust starts off-board "in-hand". Once per game, instead of moving a piece on the board, the owning player may elect to drop the Locust on any empty square. It may be used to give or parry check.
Third, in the Queen's position we have a new short-range piece: the Wallaby. The Wallaby is sort of a cross between an omni-directional Checkers/Draughts King and a limited form of Cannon. The Wallaby may make any of the four following moves:
Note that a Wallaby can only capture by replacement when leaping over a friendly piece -- you may not leap over an opposing piece to land on an opposing piece.
The following diagram shows the possible moves of the Wallaby:
The white Wallaby can make non-capturing step moves to any of the squares marked with green circles, non-capturing jumps over the white Pawn or Bishop, could leap over the white Knight to capture the black Pawn, or could leap over the black Rook to capture the black Rook.
Finally, A Pawn may promote to a Wallaby, an Edgehog, a Reflecting Bishop, a Nightrider or a Locust.
The setup for Edgehog Chess III is:
Ed Friedlander has also implemented Edgehog Chess I as an applet.
This information is based on the descriptions in Pritchard's Encyclopedia of Chess Variants and in Dickin's A Guide to Fairy Chess.
There is an implementation of Edgehog Chess for Zillions of games (the latest revision being 1.1). You can download it here: