Bruce: if I had to guess, I would suggest that it's a combination of chemical and physical, but mostly chemical. Notice in the right-hand pictures that the "moving black spot" is still present, but noticeably larger, in the dark-adapted eye. I suspect that is a physical change, analogous to our pupil widening (iris opening), while the overall darkening is a chemical change, analogous to the increased concentration of non-bleached rhodopsin as human eyes dark-adapt. But of course this is speculation. The difference in appearance of the ocellus, I'm inclined to think is just due to the lighting angle -- narrow-angle sun from upper left, versus wide-angle photoflood from center right. Note the very different specular reflection from the surface of the compound eye.
Wim: at the time, I tried searching Google with fairly specific phrases but I did not find anything relevant. I tried again today using
mantis eyes dark adaptation (which I thought I had tried before

), and got better results. On quick read,
http://jeb.biologists.org/cgi/reprint/148/1/353.pdf suggests that the effect may be due to "massive pigment shifts in the ommatidia" (contradicting my speculation above). Thanks for the link that you provided -- I will read it later today.
Ken: I gather you're talking about things like UV-darkening sunglasses? Based on very limited observation, I don't think wavelength is a big factor here. Incandescent lighting caused lightening too.
It would have been great fun to explore all this in more detail, but unfortunately the mantis escaped sometime in the middle of the second night. I was keeping it "leashed" with a piece of dental floss tied in a snug loop between the middle and rear legs. That worked great for a long while, but imagine my surprise on the third morning when I found the mantis gone, and the leash lying limp although still perfectly tied! I am fascinated to imagine what body positions the mantis had to achieve, to slip out of that leash. Houdini could not have done better!
--Rik