Working at a marine station in Australia, I caught several Abdopus aculeatus and set them up in individual tanks on a sea table. Each tank was fed seawater from an overhead pipe that fitted into a hole in the tank lid. Water flowed out of the tank through a stand-off drain fitted with a cloth filter. This type of system is nearly escape proof as long as the lid is secured. It turned out that I had more tanks than I did overhead pipes, so I connected one pipe to two tanks using a "T" fitting. As luck would have it, the connected tanks were adjacent and housed a male and a female. They would occasionally display to one another through the glass but that was the extent of their interaction. One morning I came to check the animals and found the male in the females tank mating with her. He had crawled up through his water inflow pipe, across the cross connector and down the female's supply pipe and into her aquarium. Impressive, particularly considering that the pipes were opaque and a very tight squeeze. Also, given that water was flowing from the common supply and through both arms, there wasn't any chemical signal for the male to follow. It took him several days to plan his route and execute it to engage in a tryst with the object of his desire.
Do I believe that the male planned his route to visit the female? Of course not! The behavior was fairly predictable. In a small tank with no lair, the male was constantly exploring the aquarium. The water pipe was not covered and was a tight fit, but A. aculeatus often explore tight crevices. The male was just lucky that he traveled straight across the "t" and didn't turn up and enter the main water supply. Who knows what supply line he might have come out through. Several of the aquaria housed puffers or triggers and he would have been lunch.
What really amuses me about this is that the story of the love-stricken male octopus quickly spread through the station and most people assumed that the male's jaunt was planned.
Roy