The product names used here belong to their respective trademark owners and are used here for identification and educational purposes with no trademark infringement intended.

Silicone and relatives

GE Silicone II: This is the standard to which I compare everything else. It has the least smell of any of the silicones (very slight ammonia smell), sticks well to glass but rubs off of the table rim and flows well. It does form small cracks and bubbles when drying so often I hide this by using wispy or seedy glass. The glue gets slightly cloudy when dry – I find this good for showing off the color of cathedral glass in my tables: you don't really want to see your knees under the table clearly anyway.

This is the best one to use if you don't want to asphyxiate your students or yourself and you are gluing glass to glass. It also sticks well to cement board -- use white instead of clear. Be sure to use the Window and Door version because the Kitchen and Bath version has a mildecide which yellows in sunlight. Pay attention to the "use by" date -- if it appears grainy out of the tube, don't use it as it may never cure. Removing excess is by scraping or noxious chemicals. Acetone doesn't really dissolve it, it merely spreads the film around.

GE Silicone I: stronger smell (vinegar), slightly more viscous, sticks to everything so you have to be more careful. It has fewer but larger cracks and bubbles after drying so I often use this with cathedral glass when I want less distraction. It also gets slightly cloudy when dry. It is acidic so may cause unprotected mirror edges to tarnish. You know when it has gone bad because it is has hardened in the tube.

Dow Aquarium Cement: very viscous so it is difficult to stick down without bubbles. Slightly cloudy when dry. Strong vinegar smell.

Dow Special Clear: dries very clear but is also very viscous so it is hard to stick down without bubbles. The fumes are really bad.

E-6000 (From Eclectic products. E-6800 is the UV resistant version for outdoor stuff): Probably related to rubber cement, this is the glue to use if you are gluing glass to metal etc. All the glass jewelry folks swear by (and at) it. After curing, it holds even when wet The fumes are strong so use it in a well ventilated environment.

PVA

Weldbond: a water soluble glue similar to furniture glue: starts white, dries to clear. My sample piece has been drying for 3 years and is still white in the middle of a 4 inch piece (the company said they fixed this problem but didn't offer evidence or a sample to try). It turns white when damp so you don't really want to use it in an outdoor application if you want clarity. But it doesn't seem to shrink -- after it dries you don't have any more bubbles than you started with. It works well for indoor mosaics as long as you use small pieces. And the solvent is water

Cyanoacrylate and similar formulas

These were originally made as surgical glues so they break down with time, accelerated by UV exposure. Even the cyanoacrylate “windshield repair” glue I bought at the hardware store turned cloudy after several years. Soaking in acetone can dissolve it.

Novus uses a special mix for windshield repair that is UV resistant and won't absorb water, both of which can be problems with the some of the other brands. But it is $60 or so per ounce. Good for crack repair but too expensive for the large areas I need.

Epoxy, either chemical or UV cure.

With the chemical cure, any problems are blamed on not getting the proportions right. With the UV cure, you need to be using transparent glass. These products (Hxtal, Locktite, Norland etc.; also see stone repair essay) are not generally good at filling gaps so both pieces of glass need to be smooth.

Heat Set Adhesive

Streuter's new NO Days Mosaic adhesive works on flat glass surfaces but doesn't spacefill well, even with two layers. It must be heated (with a heat gun or their special tray) to over 160 degrees F to liquify. Streuter says that in testing a double pane window with dark glass, temperature of the glass only reaches 132 degrees F on a hot day in Arazona so it shouldn't re-liquify in normal conditions. They have also done 20 year accelerated UV testing and found it stable. I'm going to try freeze-thaw testing.