These are Cartesian divers that demonstrate more "physics" than the usual diver. They work like the normal diver based on Archimedes principle. When you squeeze the bottle, the increased pressure forces more water into the diver. When the weight of the diver (including water inside) exceeds the mass of the displaced water of the diver, it sinks. Well, actually it is harder to get it to sink the first time, because surface tension keeps it up until wet. Add some surfactant (soap) to reduce this problem.
This is also Dr. A's version of Feynmann's Sprinkler, which is discussed in the book, Surely You're Joking Mr. Feynmann. Feynmann asks what happens when you put a garden sprinkler at the bottom of a swimming pool and suck water through the hose. Does it run forward or backward. Every Ph.D. physicist says it is obvious and has an answer, but the answers are all different. The sprinkler was discussed extensively in several articles in the American Journal of Physics, coming to no consensus either experimentally or theoretically. The editor then refused to accept further submissions. Enough.
In the divers shown above, the holes are drilled into the bottle cap off center like the lawn sprinkler. Does it spin the same way as a lawn sprinkler when the water rushes out and the opposite way when the water goes in? Or does the water going in hit the cap and make it go the same way as when the water exits? Or does it not move at all as some claim?
I thought Professor Moretti of Mechanical Engineering gave the an interesting answer when I showed it to him. I squeezed the bottle quickly and repetitively. He watched what happened and said, "Jets, you can blow a candle out, but you can't suck it out." Bravo! The references below give the best answer and observation that the diver does not turn when taking in water.
Thanks to Mike Lucas and the Physical Science Machine Shop for cap construction.
An elementary treatment of the reverse sprinkler, Alejandro Jenkins, Am. J. Phys. 72 1276 (2004).
A Neglected Lesson from the Cartesian Diver, Paul Kirkpatrick, Am. J. Phys. 10 160 (1942).