Take a 1M flourescent bulb, comonly seen in public buildings. Stand it upright. Have several gases pumped through it, in cycles. Ie. Hydrogen is currently filling the bulb (pink) then mercury vapour (blue) is slowly pumped in through the bottom, forcing the less dense hydrogen out of the top into a holding/gas separation tank (the tank works on gas density principles). The cycle may continue using other gasses, that make other colours. The lamp would have to cycle through the gasses in relevance to density going from least dense to most while pumping in through the bottom. Then backwards, while pumping in through the top.

So the appearance is a light that starts filling up with different colours, if the bulb was big enough, then the flowing and swirling of the gases could be observed.
-- swimr, Jun 16 2004

Color changing neon http://www.centrals...on_Rods_page_1.html
"LITEGLOW CCN2065 COLOR CHANGING NEON ROD (15", QUAD COLOR - YELLOW TO GREEN & PINK TO BLUE) " [half, Oct 04 2004, last modified Oct 05 2004]

Nice halfbaked idea! Bun for effort.

The visible white light from fluorescent bulbs is primarily due to the excitation of phosphors coating the inside of the tube, and the (UV) energy to excite the phosphors comes from glow discharge of the internal mercury vapor.

So for colo(u)r change, you may be talking about direct glow discharge from the various gases - red/orange from Neon, Bluish from Argon, etc.

Depending on applied voltage and internal pressure, interesting features like Newton's Rings (? - just did a search, couldn't find if this is the right name, but I did create some cool rings with vacuum and 20kV back in the early 1970s!) could be displayed.

Externally applied magnetic fields can also create odd/interesting effects.

Might be easier to do more of a lava-lamp design in a standard bulb, with internal incandescent light/heat source, surrounded by oils/waxes of diferent specific gravity.
-- csea, Jun 16 2004



random, halfbakery