...and that is first-hand, SWIX can verify many successful dreams...
WARNING: like many clandstine procedures this is dangerous as the equipment is not stardard lab equipment but improvised instead, meaning it is not tested (unless you test it...good idea) and so take precautions and always put safety first!!!!!
Take one wine bottle, drink wine. Sober up, or, for the more adventureous, just jump str8 into the mission... Now the bottle type must be the thick-walled type wine bottles. Most are, but bee careful the odd ones are just thin glass like spirit bottles and these will shatter(!) Get some cloth-type tape. Normal vinyl/plastic tape will not work, it stretches/goes soft when bottle gets hot and slides off bottle. So tape with cloth content will withstand heat. Take couple meters ~8mm plastic tubing, coil around bottle, beginning ~4 inches above bottom. Keep coils together, wind on all the way to the top while taping on to prevent uncoiling. Cover whole coil in bit more tape to ensure it strong enough. Take another couple meters ~8mm tubing and jam one end in 1 inch piece ~13mm tubing. This then plugs into bottle (should be tight seal). 8mm tubing goes str8 up above bottle 1 meter or so then to whatever vent/system you want to do. Bottle goes in pot of oil on element. Temp is taken up *VERY SLOWLY* and never ever above a slow bubbling of rxn in bottle. 1.5 - 4 hours is usual for hypo, 4 - 8 for I/RP, depending on size of rxn quality of reagents etc.
Whoops forgot to add that water goes thru the tubing coiled around the bottle. If you hadn't figured that out already...
WARNING: NEVER EVER REACT I/P OR I/HYPO IN THE BOTTLE! VERY IMPORTANT! DO THIS IN PYREX FIRST! HEAT FROM RXN WILL SHATTER BOTTLE!!! (Learnt first-hand also...

)
Afterwards dismantle and dispose of the bottle. SWIX used to like the fact that SWIX could walk into supermarket, spend $200 on jars dH2O matches wine paper towels etc etc and when finished whole lot goes into rubbish and no evidence (as long as rubbish carefully disposed of of course).
One last point: testing glass strength is pretty easy. Build it, put it on with just water in it and boil the water vigourously in the oil bath. If it doesn't develop cracks (or just outright shatter) you got a winner. Apart from the time I/RP was mixed and the bottom dropped out, SWIX has had 100% success.