Glass Corrosion

What is the wetting fluid, and at what temperature and pressure?

Have you determined that corrosion has occurred by measuring some loss in weight of the apparatus, or by detecting the presence of leachates? If, after a run, the apparatus looks 'scuzzy', it could be that reactants/products are adhering (H-bonded?) to the glass, and the underlying glass is actually not being affected. In general, acids tend to leach the alkaline metal 'fluxes' out of glass, whereas alkalai (especially OH-) attacks the Si-O bonds (true corrosion). Both of these will result in grungy-looking glass, but acid attacked glass tends to appear frosted when dry, and alkalai attacked glass tends to look greasy. Note though, that the acid-frosted glass will appear similar to undamaged but encrusted glass, except that acid attack is subtractive and encrusting is additive to your system. Might this be possible to determine by careful weighing before and after a run? If acid attack is occuring, consider using a 'harder', flux-free glass: fused quartz. If alkalai attack is the mechanism, you may be at a dead end. The only idea that pops to mind is to apply some sort of resistant coating to the apparatus, or use plastic instead of glass.