A direct proof of the non-existence of a \(\mathrm{MOL}(6)\) (Q6168183): Difference between revisions

A celebrated combinatorial result is the nonexistence of two orthogonal Latin squares of order 6. This was conjectured by Euler in terms of his famous 36 officers problem. It was possibly proven by Clausen in 1842 (unpublished) and certainly by \textit{G. Tarry} [C. R. Assoc. Fr. Av. Sci., Paris 29, No. 2, 170--203 (1900; JFM 32.0219.04)]. Since then, a number of other proofs have been found. This paper provides yet another. Tarry's original proof was by exhaustion, but some of the later proofs introduced some interesting ideas and came closer to providing an explanation, rather than just a proof. The present proof does not really introduce any important new ideas, or do much to explain the non-existence of a solution to the 36 officer problem. It is sensibly methodical, but ultimately a case bash. And at over 15 pages, it cannot claim to be short either. The main 3 cases that it divides the problem into are (1) Latin squares that contain two rows that contain 3 intercalates (subsquares of order two) (2) Latin squares that contain no intercalates and (3) All other Latin squares. It does not rely on generating a catalogue of isotopy class representatives of Latin squares of order 6.