I've had several letters about chemical colors. One chemist friend points out that for wild colors, the indoles are the way to go. I haven't done too much indole chemistry (which is OK, since they can reek), but he's right about that. You can get all sorts of pinks and purples as side products. The classical indigo pigment is a derivative of this structural class, so it makes sense. Of course, none of the colorful ones are going to be drugs, most likely, since they have lots of weirdly conjugated bonds that make them both colorful and reactive. Not something you'd fork over for at the pharmacy.

Another letter points out that I misstated Carl Sagan's term for the organic goo found in the outer planets (and in my fume hood at work.) It's "tholin", not "thiolin," which removes my objection completely. And it's what I get for doing these things from memory, when I could reach three feet over to the bookshelf and check it out.

The same reader mentions that recent calculations have shown that if someone manages to make 1,3-difluoroazulene, the color should shift from blue to emerald green. That would be worth seeing, but it's still not enough to get me to do serious fluorine chemistry. You can mess things up very thoroughly with some of those fluorination reagents (and I, unfortunately, can tell you what xenon difluoride smells like, so I have no desire to investigate the other possibilities!) For the record, it's not really a bad smell, just not something that should necessarily be smelled. . .

That reminds me of an entry in the Merck Index. I don't have a copy at hand (this time!), but I think it's for something hideous like fluorine oxide. The entry notes "extremely poisonous. . .attacks lungs even in traces. . .peculiar smell." I've always thought that rather evocative. That last observation sounds like someone's last words, frankly.