In the moonlight other retinal cells called rods do most of the work. Rods detect relative lightness and darkness, but they are completely color blind.
Moonlight is simply the white sunlight reflecting off the gray surface of the moon. There is nothing in this interaction to give light bluish or greenish quality. In fact, scientific instruments have shown that the light of the moon is slightly redder in color than the direct sunlight.
These facts suggest an added mystery at the heart of how we as artists chose to depict the moonlight in paints. If the moon is just light gray in color, and if it is close to the minimum threshold of our color receptors anyway, so why so many artists paint bluish or greenish moon? Do we really see it that way? Is it a kind of illusion, or perhaps is it just an artistic convention?
look at some paintings by master painters of moonlight. As you watch, think of your own color vision in the night, and ask yourself which of the paintings convey best your own experience.
This is a picture of J.M.W. Turner. It is quite gray, with just a hint of warm color around the moon. Note that there is not much detail in the shadow zone. All you really see clearly are the silhouettes of sailing and the boat on the water. 
Russian seascape painter Aivazovsky painted this night scene lit by a golden moon. The sky, water, and all shadows fall in blue-green tones. It does not show many details, and it stops well short of the black in the shadows.
This relief darks was also a characteristic nocturnal Remington. The cast shadow to the left of the pony's nose is composed of umbers and dull green. Bright light and shadows enliven the dark. Except for the light saddle blanket, Remington has left most vague and indefinite edges, especially on the donkey on the right.
This famous night in Whistler Battersea Bridge uses a fairly saturated blue-green color, especially in the water and in the figure silhouette. The detail is blurred everywhere, even in areas where the bridge appears in the sky, setting up for small bursts of light in the distance.
One reason to soften the edges is that we depend on cones for fine discrimination edges. Unfortunately, the cones are located on the fovea, the central point of vision, and with them offline in the dark, we can not adjust the small details.
If you take a book or a newspaper outdoors in the moonlight, you can see that there is written on the page, and you might be able to read newspapers or other large characters, especially when you look around your peripheral vision. But reading the regular size text is almost impossible. When you look directly at the words, blind spots get in the way.
I said earlier that our cones are sentence running in the moonlight. In fact, contrary to what some authorities have said, the cones of most people can do basic color judgments in the light of the full moon. But how much color variation can you really see?
Maxfield Parrish made this clear moon scene with a little color saturation. He painted the bright yellow moon, the red cup on the barn, the deep blue of the sky and the orange color of the shady side of the house. Did he really see these colors in moonlight, or he invents for pictorial effect? Too bad he is not here to ask.
direct outdoor painting is virtually impossible in the moonlight. Each artist has to work from memory and imagination. We can try to convey our actual optical sensations, but we're not scientists. Each of us also trying to make a statement subjective aesthetic meant to evoke a particular mood or emotion. Any painting by moonlight is an attempt to translate a "rod of experience" in a "cone of experience," an image that will be visible in a well lit environment.
is how you can test your cones actually respond to color in the moonlight. Paint a set of matching, unmarked separate swatch or find a construction paper about the same value. Take the full moon (Tuesday) and let your eyes set (about 30 minutes it takes). shuffle the cards, and while you're still outside, mark on the back what colors you think they are. 
I used Photoshop to manipulate a photo swatches (actually shot in daylight) to simulate how they appeared to me under the full moon: dulling, darkening and blur both Jeanette and I could easily identify. the family of the basic shade of each color chart. But beyond this basic classification, we did not know, and the shade of gray we two confused. When I looked the same samples in the light of a much smaller half-moon or a shadow of the moon, I found my cones went subthreshold and closed completely, and swatches became completely monochromatic.
Although the stems of the eye can not actually see the color, scientists have shown that they are the most sensitive to greenish wavelengths of light. Accordingly blue-green colors appear lighter in tone in dark conditions. There's a name for it: Purkinje Shift. It is a different phenomenon and often mistaken perception of the moon blue.
You can demonstrate Purkinje Shift by comparing a sample red and green that begin within the same value. If you take them outdoors in the moonlight, a greenish seem much lighter in tone. Many observers have noted that red roses look black in the moonlight.
If you scroll back up to my Photoshopped version of the swatches in the moonlight, you can see that I have adjusted the values to simulate the way the red and green seemed due Purkinje shift.
here, Remington shows a scene with the Indians in the moonlight. We see their skin tones and some clear red buttons in their costumes. All along, the edges are much sharper than the other paint.
That night the old Cincinatti by the contemporary artist John Stobart has a distinctly bluish cast. It introduces a lot more detail than we have seen in other examples, reminiscent of the film "day-for-night" shoots in old westerns. You can even read the name "Bonanza" the shade side of the ship. in addition to the moon, there is a secondary source of yellow-orange light. in this case, one could say that the blue cast to the image can be a complementary color induced by the opposition color of the lamp.
Atkinson Grimshaw was famous for his studies of poetic clear moon. Here, the masses of shadow on the left are quite soft and impenetrable, but bricks and branches appear very clearly. the moonlight on the road is an intense yellow-orange, assuming that the reproduction is accurate. the shape of the light spot patch lovers standing in silhouette on the left. 
Russian realistic landscape Ivan Shishkin painted this haunting image of a winter night in the wild north. The snow in the moonlight is relatively bright, with a soft halo along the edge to the left, not yellowish. The gradates cast shadows in tone, more light it catches more sky filling and bounced light. There are a few details in the form of the tree, but he kept the foreground and background description to a minimum.
So to go back to the earlier question, why do we see the light of the blue moon?
Saad Khan and Sumanta N. Pattanaik, University of Central Florida have proposed that the blue color is a perceptual illusion, caused by an overflow of neuronal activity of the rods to the adjacent cones.
a small bridge between synaptic active rods and cones inactive key off the blue receptors in the cone, rather like an insomniac turning over in bed and wake his sleeping wife. The influence of activity on stem cones adjacent turns the brain into thinking that we see light blue, even though we're not really
As the authors say :. "We hypothesize that cells mainly stem synapse on the S -Cone (cells sensitive cones in the bluish light) circuits resulting in the visual cortex receiving a blue hue"
thus, moonlight is not blue. our eyes are just playing tricks.
Unfortunately, this hypothesis remains untested tantalyzing. I contacted Dr. Khan and he said that because of other projects, he has not had time to prove the hypothesis under controlled conditions. I hope it can shed more light in any color on this elusive subject.
Until then, the moon remains a mystery to the point of meeting art and science.
Further reading
- Khan's review article and in Pattanaik Journal of Vision , 04. Link
- Related discussion on the NASA Web site .. Link
- "the Eye and Night Vision" of the American Optometric Association. Link.
- More information on Remington night the blog of David Apatoff Link
Tomorrow. Elegant Graphics
