Ever had to run a reaction with a volatile gas? It's not a very common thing to have to do, but every once in a while, it needs to be done.
Assuming the volatile gas is coming from a lecture bottle, there are a few ways to do this:
Bubbling into solvent: Add solvent into a reaction flask (with stirbar?) and pre-weigh. Bubble gas into stirring solvent (cooled?), stop when the weight matches the desired amount. I've been concerned about suckback into lecture bottles (although I suspect I'm very paranoid, so there are other ways to skin this cat.)
Just doing it: From the brass adapter (that you can buy special from Aldrich!) that's screwed onto the lecture bottle fitting, attach a tube and direct it into a pre-chilled tared round-bottom flask. When you think you're done adding, turn down the gas and quickly weigh the flask.
The methods above work fine, but you run the risk of allowing a lot of evaporation of your gas to happen. So, when you move to a higher scale (>20 grams of condensed liquid), you might want to try a coldtrap. Assuming that you have one (see 1st picture below), it's pretty self-explanatory: just a glass coldfinger with a place for dry ice/acetone to cool down the incoming gas, a gas inlet above and a 24/40 ground glass joint at the bottom.
You prepare your lecture bottle pretty mundanely (2nd picture): screw your brass fitting onto the inlet and attach whatever tubing that you prefer. I like stiff plastic tubing, but that's just an affectation more than anything. Obviously, 1/4 inch Tygon is best for good seals between tubing and fittings. If you have something that's particularly dangerous (it is, after all, a pressurized gas), you should probably clamp that thing down. While it's not pictured below, it's probably still a good diea.
When I was running these reactions, I was always concerned about losing gas. Using the "just do it" method, I think I was losing up to 30% of the gas I was trying to condense. The cold trap method dropped that significantly (<5%).
An NSF funded Project