See also D. E. Nilsson and S. Pelger's article, "A pessimistic estimate of the time required for an eye to evolve"
(Proceedings of the Royal Society B, 1994, v.. 256, pp. 53-58).
In his recent book, River Out of Eden (Basic Books, 1995), Richard Dawkins points out how Nilsson and Pelger set up a computer model of evolving eyes to determine if a smooth gradient of change exists from a pigmented eye spot to the camera eye with a lens and cornea, and how long it would take such a transformation to occur. They employed pessimistic figures for the amounts of change possible per generation -- giving their model only 50% "heritability" (many human traits are over 50% inheritable), and chose pessimistic values for the coefficient of variation (how much variation there typically is in a population). And they determined that Darwinian evolution could produce a good camera eye in less than a half a million years! That's a mere "blink of the eye" in geologic time!
Since an eye's efficiency can be easily measured using elementary optics, their computer simulation had more validity than, say, trying to measure how subtle anatomical changes increased the efficiency of a cheetah's speed and agility.
"Nilsson and Pelger began with a flat retina atop a flat pigment layer and surmounted by a flat, protective transparent layer. The transparent layer was allowed to undergo localized random mutations of its refractive index. They then let the model deform itself at random, constrained only by the requirement that any change must be small and must be an improvement on what went before. The results were swift and decisive ... leading unhesitatingly from the flat beginning through a shallow indentation to a steadily deepening cup. The transparent layer thickened to fill the cup and smoothly bulged its outer surface in a curve [the cornea]. And then, almost like a conjuring trick, a portion of this transparent filling condensed into a local, spherical subregion of higher refractive index [a lens]." -- Dawkins, pp. 80-81
And the lens that formed was not of a uniform refractive index, but was "graded," just like real eyes, with the highest refractive index near the center of the lens! And it was graded according to the optimum ratio for vision, known as "Mattiessen's ratio."
I should add that Nilsson and Pelger's computer simulation never produced an eye that combined the focus of two lenses -- one placed directly behind the other -- lenses that could slide toward and away from each other to produce added magnification and "close-ups" of small objects and far away objects, as in a "zoom camera." Instead, the best "zoom" available to us humans is to bring the newsprint closer to our eyes! I guess the "Biblical Creator" in his infinite wisdom could not design eyes any better than natural selection could. However, robots of the future will undoubtedly have such "extra" design features added by their human creators.
Other recent articles, like Gould's, mentioned above, have pointed out how a common genetic key triggers the development of eyes of vastly different construction in animals as varied as flies and mice (in vertebrates and invertebrates). So, all eyes may originate from a common ancestor that evolved this genetic trigger. See for instance, Peter Monagham's article, "Revelations from Fruit Flies" (Chronicle of Higher Education A8-A9, May 26, 1995). And also see Carol Yaesuk Yoon's article, "The Wizard of Eyes: Evolution Creates Novelties by Varying the Same Old Tricks" (New York Times, Nov- 1, pp. C1, C11).