I was reading the latest issue of Scientific American last weekend — a familiar and comfortable pursuit. I've read this magazine regularly since I was in elementary school, and much of what I know about science comes from it.
There's a very interesting article on vision. It's long been known that the human eye is in constant motion, even when we're attempting to stare straight at one point. People have long suspected that these miniature tremors are essential for vision, and this article cites experiments where spikes from optical nerves diminish and cease when microsaccades are artificially suppressed. (You'll have to read the article to find out how this is done.)
As with most such studies, all the cited examples involve photopic (daytime) vision. But when I saw a diagram with both rods and cones labeled, I suddenly realized that it might be relevant to scotopic (night) vision as well. In fact, it's probably even more important, because night vision is famously sensitive to motion.
That's when I thought of a process that I usually describe as "scanning the field carefully with averted vision." It's a funny activity — I always feel as though I'm moving my eye in an unusual and slightly unnatural way. It requires real muscle control, like threading a needle. I wonder if this is partly a question of putting those microsaccades to use? They're semi-voluntary — can't be suppressed entirely, but can be diminished by conscious effort. Maybe when I'm stargazing I'm actually boosting the microsaccades.
I'd always thought that learning to see faint objects was all a question of mental training. It never before occurred to me that it may also have a simple physiological component, like strengthening or training the muscles that move the eye.