Isaac Newton himself tried pig bladders to support mirrors. Back in the early 70's the old Telescope Making Techniques magazine, which was a short lived (four issues) followup to the Maksutov Club circulars, had an article about pneumatic support.
A passive bladder is too simple. At the zenith the mirror would sag longitudinally and at the horizon it would rise. There is only one elevation where the collimation points wouldn't distort the mirror. Pneumatic edge support is problematic because you need pressure only at the lower side of the mirror, not all the way around as your 16 inch inner tube would provide. There are telescopes with liquid (mercury) edge support, which can be designed to produce the right effect.
I have a 20 inch Dob whose mirror is supported by air. It has a dynamic system that adjusts the pressure to perfectly float the mirror as the telescope's elevation changes. Here is a rough description:
My mirror is supported by a ribbed vinyl floor mat with the ribs on the mirror side so it can cool. The mat is glued at its edge, about 1/8 inch beyond the mirror's edge, to a 1/16 inch thick PVC ring which is in turn super-glued to a smooth, varnished birch plywood board. There are three collimation pads under the mat at the edge of the mirror; they're actually inward-going areas of the PVC ring. The mirror clips are directly over the pads. The whole board moves in collimation.
The mirror is supported by air pressurized by a smaller piston, which is a three pound scuba diving weight bolted to a round piece of plywood, which presses on a balloon (actually a whoopie cushion, flat and round, about six inches diameter when uninflated) which has a rubber stopper in it and some aquarium tubing connecting it to the mirror cell. The small piston is hinged on a board so that it hangs in the same plane as the mirror, in a sub cell below the mirror board. The diameter of the small piston is about 5 inches (I can't remember exactly), which with the three pound weight produces just enough pressure to float the 20 inch mirror. A tee valve in the tubing line goes to a rubber sphygnomanometer bulb which is used to initially pump the small piston to the middle of its range. After that it is automatic: at the zenith the system pressurizes to about four inches of water (2 inches of pyrex at specific gravity of about 2), and at the horizon the pressure drops to zero.
There are lots of details but those are the basics. I have a conventional sling for edge support, by the way. Here are some points to consider:
--the volume of air under the mirror must be kept small so that you won't need to produce a lot of compensating air volume.
--atmospheric pressure varies a lot. The 4 inches (water) of pressure acts on top of 30-some feet of atmospheric pressure. The small piston has to have enough capacity to take care of weather fronts moving through and to accommodate temperature-related volume changes.
--a closed volume of air is a good insulator so you should use something like a ribbed mat so that the back of the mirror can cool. I made a plenum and use a fan to draw air through all the rib channels so the mirror is cooled very effectively.
--with collimation pads at the edge you can put the mirror clips directly over the pads and hold the mirror firmly with true "hard points", without distortion. Because the mirror is fixed firmly between the collimation pads and clips, the collimation stays very good, and the telescope never needs "thumping" to make the mirror settle.
--because the mirror has only three collimation supports the compensation pressure has to be quite exact because the mirror would distort if it has to "work" against those supports or the mirror clips. The actual weight of the mirror must be supported by the action of the smaller piston. I made a rig with a differential manometer (loop of tubing with water in it), with the mirror on one side and the compensating piston on the other, and trimmed the diameter of the piston's disk to equalize the pressures.
--note that if the mirror is very large and the collimation points are at the edge, then you don't want to transport the mirror lying on these points. The mirror sling should support the mirror when transporting (but that's the normal telescope orientation during transport anyway).
--the support is very good, I think better than a many-point mechanical flotation cell, but it isn't perfect. The pneumatic cell provides one pressure (at any elevation) but the thinner center of the mirror doesn't need as much pressure to support it as the thicker edge. So the center of the mirror is pushed out. It is negligible for all but the thinnest mirrors and interestingly the effect is to undercorrect the mirror, which is "good" because mirrors become slightly overcorrected as they cool.
All in all, it works well but it probably isn't worth all the trouble unless you have a very large mirror which is difficult to support mechanically. The only moving part is the small piston. Mine has a six inch one-piece plastic piano hinge that can't bind. There are other ways of producing compensating pressure. You can put a circle of large-diameter vinyl tubing inside a Dob's altitude trunnion, fill it partly with antifreeze so that it acts as a manometer that shows a 4 inch differential at all altitudes. Getting all the volumes and dimensions just right would be tricky but it would look really cool if it worked. In a fixed observatory you could have an aquarium pump provide the pressure--even a small pump could support a very large mirror--via a mechanical or electrical regulator.
Aart M. Olsen aart@uiuc.edu 217-333-7467 College of Veterinary Medicine Univ of Illinois at Urbana-Champaign