On embeddings of snarks in the torus (Q952646)

In the paper, there is a condition on cubic graphs \(G_1\) and \(G_2\) , implying that a dot product \(G_1\cdot G_2\) has an embedding in the torus. Using this condition, it is proved that for every positive \(n\) there is a dot product of \(n\) copies of the Petersen graph, embeddable in the torus. This disproves a conjecture of Watkins and Tinsley and answers a question by Mohar.