Perk (perchloroethylene) and other chlorinated solvents were once the fluid in common glitter lamps. All the vintage French, Italian, UK etc. lamps, as well as the 80s Chinese import glitters like YAPS and TLC Hollywood, used this. Trouble is, it's nasty stuff. It gives off chlorine vapor, and can eat the silvering off the glitter, which was commonly metallic-coated Mylar flakes. The liquid used in the Lava brand GemLites and Wizards was similarly nasty, though SteveSF works with Wizard liquid and can tell you more about it.

I have found no safe liquid and glitter which work. I've done a lot of tests with diluted glycerin, propylene glycol, etc. and nothing seems to hold the glitter, which (depending on the type used) either floats to the top and stays, or sinks to the bottom and stays.

My way of "making" glitter lamps is to buy the cheap new imports - my favorites always have silver, rather than transparent irridescent, glitter - open them, and then fill my bottles with the commercial solution. If you can get hold of extremely fine aluminum leaf (think gold leaf - microscopic thickness - but aluminum is so much lighter than gold) you might try that, shredded, in glycerin. I've tried both real and imitation gold leaf and silver leaf, the imitations are too heavy also, but I believe the "slow glitters" (Florence Art Co., MasterCrafters, Lava brand GlitterLite, and Crestworth slow GlitterLite) used this mix.

There is a Crestworth patent (no. 1,232,311) for the "slow" glitters, using a mixture of polyethylene glycol and propylene glycol for the liquid phase, with aluminium coated plastic flakes for the glitter.
See link;
http://v3.espacenet.com/publicationDetails/biblio?KC=A&date=19710519&NR=1232311A&DB=EPODOC&locale=en_EP&CC=GB&FT=D

There's some interesting info from Mike Firth on this page with experiments making glitter lamps and the solution and materials experimented with... http://www.anet-dfw.com/mikefirth/lavasubs.htm