It was an undergradute senior thesis, so please excuse the amateurishness of the science...
We had small Sepia in aquaria which were mostly bisected by a clear plastic wall. There was enough space on one side of the aquarium for the cuttlefish to easily get around the wall, but they didn't seem to do so very often. The amount of space on the two sides was unequal--one side (the side where the cuttles spent most of their time) was probably twice as large.
On the smaller side, there was a tube which led into the water, about 1 inch diameter.
In each trial of the experiment, we would start by playing a sound in the cuttlefish aquarium through a submersible loudspeaker. I don't remember the exact frequency or anything--might have been 1000 HZ, probably less. In any case, it would go on and off for about 15 seconds, after which a small crab would be dropped into the tube. We would time how long it took the cuttlefish to catch the crab after the sound stopped, with the theory that, if they could learn to associate the sound with the food, they would go around the wall and be waiting when the sound stopped, and catch it very quickly.
We didn't find, in general, that they got any faster at catching the crab, even after several weeks of daily training. They did, however, certainly react to the sound, and we got some interetsing data on what frequencies and intensities they responded to. With loud, low sounds, they show rapid color change and escape behaviors. Yes, I spent a quarter scaring the hell out of cuttlefish.
In any case, it was a long time ago. In the unlikely event that anyone wants specifics, I could dredge them out of the dark corners of my hard drive...the study was not published.
