The same year, Heawood proposed a conjecture which gives a lower bound for the number of colors that are necessary for graph coloring on a surface of a given genus. It was proved in 1968 with one exception : the case of the Klein bottle.

Response to Mathematica on X