Sunday 22 April 2018

Magic Four-Giraffe Tour From 1887

On 9 October 2013 I reported here my construction of a Magic Two-Giraffe Tour, i.e. consisting of two sections of {1,4} Giraffe moves joined by two rook moves. In that tour the diagonals add to 272 and 248 which together sum to 520 which is twice the magic constant of the ranks and files.

While studying the cryptotours published in Le Siecle in the column "Un Probleme Du Jour" edited by A. Feisthamel, from 1876 to 1894, I have found this much earlier work on the same subject. This uses four Giraffe paths connected by four rook moves. In this tour the diagonals add to the magic constant 260 as well as the ranks and files.

Diagonally Magic Four-Giraffe Tour by A. E. Reuss of Strasbourg
Problem 3221, Le Siecle 5 March 1887, solution 12 March 1887.

01 25 09 23 42 56 40 64 
43 49 39 63 02 26 16 22 
03 27 15 21 44 50 38 62 
45 51 37 61 04 28 14 20 
32 08 24 10 55 41 57 33 
54 48 58 34 31 07 17 11 
30 06 18 12 53 47 59 35 
52 46 60 36 29 05 19 13 

The rook moves are 16-17, 32-33, 48-49 and the closure move 64-1. The tour is symmetric about the vertical axis, the ranks consisting of complementary numbers adding to 65, but is not quite symmetric about the horizontal axis.

Naturally I wonder whether Reuss constructed others of this type, or is this just a one-off? Knight tours that he published in the same column were under the pen-name of "X a Belfort".

Wednesday 11 April 2018

Sources of Magic Knight Tours: Update

Two minor updates to the list of sources of magic knight tours of the standard chessboard have come to light from my researches in Le Siecle.

Tour 00i by an Unknown composer appears as problem 772 on 25 April 1979 with solution on 2 May 1879. The word puzzle used with it is given as by M. Jacquemin-Molez but the composer of the design is not identified. The column is signed as by A. Feisthamel as usual, but if by him the designer would probably be cited as M. A. F. so I think we have to classify it as still Unknown and from an earlier source. The 1879 date is one or two years earlier than the 1880/81 date given by Murray.

Tour 00f by Palamede (the pseudonym of Count Ligondes) appears as problem 2134 on 31 August 1883 with solution on 7 September 1883. There does not seem to be any mention of its cyclic properties.