I was just reading this patent application: http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220070159925%22.PGNR.&OS=DN/20070159925&RS=DN/20070159925 (patent application 20070159925, by Chryssostomos)
It describes a new (?) kind of speaker with no moving parts, in which the liquid medium itself produces the sound: you put a couple of magnets close together across a tube filled with saltwater, then run a current across the tube at right angles to the magnetic field, and it pumps the saltwater through the tube (electrolyzing it a bit). You can pump it back and forth at more than 100kHz, and the patent claims, "In one embodiment the device 10 generates an acoustic signal having an amplitude of approximately 150 dB referenced to 1 .mu.Pa at a 1 meter distance using 0.5 Tesla magnets with a 10 ampere, 4 volt rms electrical drive signal." A micropascal is apparently the normal reference for dB SPL in water; 20 micropascals is the normal reference in air, because it's around the threshold of human hearing. So those 150dB are similar to 150 normal audio dB. I thought half a tesla was an unreasonably large magnetic field, since MRI machines have gigantic superconducting coils to make 1.5 to 4 teslas, but Wikipedia tells me that loudspeakers commonly have a 1-tesla field across the coil, and neodymium-iron-boron magnets have strengths around 1.25 teslas. Here I consider some other possible uses for similar devices. Of course, if you replace the liquid by a chunk of metal, you have a normal dynamic speaker, with a single turn of wire. So I think you can use this device as a no-moving-parts microphone as well. If you drive it with DC, you have, in a sense, a linear homopolar motor or rail gun. But you probably can't do that for very long with salt water before the electrodes are coated in electrolysis deposits. (This might also turn out to be a problem with Chryssostomos's application; I'm not sure.) Apparently when you do this with a plasma as the working fluid, the result is called a "magnetoplasmadynamic thruster", and it's a possible future alternative to ion engines; there have been some successful space trials. There are other conductive liquids that aren't ionic: mercury and molten solder, for example. I suspect it may be possible to use these liquids in this application with higher currents than are possible with saltwater, and they may damage certain kinds of electrodes less. (They will tend to dissolve other metals eventually, I think.) In particular, you could use this technique to make an inkjet of molten solder, and you could probably accelerate it considerably by directing a pulse of solder into a sort of nozzle --- a tube of decreasing diameter which would result in some of the pulse spurting out the end at high speed, containing most of the energy of the pulse, while the rest stops in the tube. (You'll need some way to get it out of there for the next pulse.) The nozzle is, in effect, a way to increase the current density. I have tried to figure out how you could use this effect to cause a bunch of fluid to converge toward a point without further direction, but that requires both the magnetic field and the electric current to remain tangent to the surface of a sphere over some substantial part of its area. I don't know how that is possible. If it is, you could get the nozzle effect without having an actual nozzle. I think you can approximate this by having a concave meniscus, which will force the current to curve along the surface, and curving the magnetic field downward by making it more intense inside of the void of the meniscus.
