Down there, the legal brightness for streetlight illumination exceeds three candelas per square yard, which is the threshold above which the human retina's cone-shaped cells operate. This yields constant "photopic vision," which lets people see sharply and in color.
But around here, when you leave Kingston or your local village, nothing's that bright after nightfall. So we only get to use our rod-shaped cells after nightfall, which bestow scotopic vision except when indoors. Scotopic kicks in when things are dim. Although our eyes have 100 million rod cells, compared with our eight million cones, the greater numbers confer no benefits in sharpness. Using rods is an awful way to perceive the world: Its sole advantage is dim-light usefulness.
First off, rods are colorblind. Next, there's not a single rod lurking in the middle one degree of vision; they're most densely packed 15 to 20 degrees from dead center. So in low-light situations when cones no longer operate, we suffer a one-degree blind spot at the very center of vision: twice the size of the moon. (There's also a second, better-known blind spot present even in bright light, caused by the optic nerve. This one's not located centrally, and we don't usually notice it: If an object is hidden at the blind spot of one eye, it will be seen by the other.) Anyway, to carry out observations using rods most clearly, you need to look about 15 degrees away from the object of interest, while nonetheless keeping your attention on it.
Another quirk of rods is a total blindness to red objects. It's not that they're merely seen as grey, like all other colors; rather, they vanish altogether!
Rods are also very slow-reacting, which is why it takes so long to acquire night sensitivity. When you first switch off your bedroom lights, you probably see nothing at all. After a few minutes, things in the room become obvious. Maximum sensitivity is reached in about 20 minutes.
We've saved the worst for last. On top of all these failings, scotopic vision only delivers 20/200 acuity - ten times less sharp than photopic vision. You've always sensed the truth of this. Sharp details (like the creases in those clothes that you've tossed onto a chair at night), which are so obvious when the lights are on, now become a blur in the dim light. We're so accustomed to it that we probably assume that dimness has something to do with vagueness. Nope; it's those rods.
This is why beginners who buy telescopes are sometimes appalled at how few details appear on galaxies and nebulae, on top of them being colorless - a big contrast in both departments from the exquisite photos in magazines. Indeed, that's a primary purpose of astrophotography. It's not just to record and preserve the image; it's to bring out stuff that the human eye would simply never see - not even through the world's largest telescopes.
We do see about 50 stars in color. Those brighter than magnitude 3.25 can excite our cone cells. All stars emit red, green and blue light - light's primary colors; when combined they create the sensation of white. So all stars mostly look white. If the star is unusually hot and has an excess of blue, this mixes with the white to produce a pastel blue-white - like Rigel, Orion's foot-star. An excess of red mixes with the white to make a star pastel orange - like Betelgeuse, Orion's shoulder. Thus, night colors, even when they appear, are never saturated, but merely pastels.
Finally, a full moon gives enough light to get the cones going slightly, while rods are still operating. This is called mesopic vision. Here, the cones weakly operate only at their place of peak sensitivity, which happens to be blue-green. That's why the natural world will appear that color under this week's full moon.
Suddenly, the night makes sense.