wow, cool -- i guess all mercury, huh? i suppose there aren't that many things that glow as bright as mercury, and if it gives you R, G and B, tinting the glass is a totally reasonable way to get colors. I do wonder if you can only get "cooler" colors since the mercury lines are so narrow. I don't mean bluer colors, but am thinking of the difference between a doped CFL and a basic mercury tube.

Mercury can account for the blue and green peaks, but not the red peaks. So all the tubes (except maybe the red one) have mercury. The red, pink, and brown tubes must include europium to produce the red lines at 580, 587, 593, 598, 611, and 630 nm.

So it appears that, like most lighting tubes, these contain a filler gas like argon with a little mercury in it. When a charge is applied, the excited mercury atoms release mostly UV light and a little visible light (the blue and green peaks at 436, 546, 577, and 579 nm). The red, pink, and brown tubes are lined on the inside with a europium compound which fluoresces orange and red when UV light excites it. The green and blue tubes do not appear to include europium.

The red tube shows no peaks from mercury, maybe because a red filter in the glass absorbs those wavelengths. There is probably mercury in there to produce the UV radiation which excites the europium.

The pink, green, and blue tubes have broad smears around particular mercury emission lines. I don't know how this is done. It might be a fluorescing phosphor, but it is not emitting discreet peaks the way most rare earths do. I don't think colored glass would have this effect. Mercury does not produce light at all of those wavelengths, so some other element must be producing it.

I have to remember to ask to see what the sign looks like when it is turned off.