second case chance is not required to work the
miracle it would have to perform in the first. Not only is the number of
resemblances to be added somewhat reduced, but I can also understand
better how each could be preserved and added to the others; for the
elementary variation is now considerable enough to be an advantage to
the living being, and so to lend itself to the play of selection. But
here there arises another problem, no less formidable, viz., how do all
the parts of the visual apparatus, suddenly changed, remain so well
coordinated that the eye continues to exercise its function? For the
change of one part alone will make vision impossible, unless this change
is absolutely infinitesimal. The parts must then all change at once,
each consulting the others. I agree that a great number of uncoordinated
variations may indeed have arisen in less fortunate individuals, that
natural selection may have eliminated these, and that only the
combination fit to endure, capable of preserving and improving vision,
has survived. Still, this combination had to be produced. And, supposing
chance to have granted this favor once, can we admit that it repeats the
self-same favor in the course of the history of a species, so as to give
rise, every time, all at once, to new complications marvelously
regulated with reference to each other, and so related to former
complications as to go further on in the same direction? How,
especially, can we suppose that by a series of mere "accidents" these
sudden variations occur, the same, in the same order,--involving in each
case a perfect harmony of elements more and more numerous and
complex--along two independent lines of evolution?

The law of correlation will be invoked, of course; Darwin himself
appealed to it.[30] It will be alleged that a change is not localized in
a single point of the organism, but has its necessary recoil on other
points. The examples cited by Darwin remain classic: white cats with
blue eyes are generally deaf; hairless dogs have imperfect dentition,
etc.--Granted; but let us not play now on the word "correlation." A
collective whole of _solidary_ changes is one thing, a system of
_complementary_ changes--changes so coordinated as to keep up and even
improve the functioning of an organ under more complicated
conditions--is another. That an anomaly of the pilous system should be
accompanied by an anomaly of dentition is quite conceivable without our
having to call f