An age-old objection to the concept of evolution is that of the blind watchmaker, and it goes like this. If a man were to find a working watch in an apparently abandoned place, far from civilization, which would be more probable: that the elements all happened to combine perfectly in the wind and heat and rain, such that a watch resulted by mere happenstance--or that another individual who had purchased the watch from an intelligent designer had been in that same place before, and had simply lost it?
It's popularly believed that this argument has been discredited, but I've yet to hear the actual counter-argument to refute it.
Darwin himself wrote of his theory of evolution, "If it could be demonstrated that any complex organ existed which could not possibly have been formed by numerous, successive, slight modifications, my theory would absolutely break down." In other words, any organism, or component of an organism, that was irreducibly complex would discredit his theory. He could say this in his day, because at the time, biochemistry was entirely unknown. Author Michael Behe wrote in "Darwin's Black Box" that it was once believed that insects arose spontaneously from dung and spoiled food, because if left to themselves for any length of time, one might go away, return, and find organic material infested with them. When small organisms were assumed to be very simple, this seemed believable. Now we know better. Even the simplest cell is unbelievably complex--akin to an incredibly efficient city in which each citizen knows and performs his job. These jobs include protein synthesis and breakdown, energy production, repair, and communication within the cell and with the outside world, not to mention complete replication of itself. The concept of irreducible complexity, as Behe defines it, is to first determine the function of the system and all the system's components, and then to determine if all of those components are required for its function. If so, then by definition, that system could not have come about by gradual changes. If it did, natural selection would have no reason to select each iteration on the way to functionality, because each in-between step would be at best, useless, and at worst, fatal.
The classic example of the former is the eye. Earlier iterations of this incredibly complex system could not see, and would thus be useless. Therefore, natural selection would have no "reason" to pass on the non-functional, half-formed system to future generations. An example of the latter is the clotting cascade: an intricate internal and external feedback system allows blood to clot without a runaway clotting process that might solidify all the blood in the body at once. If the system did not work at all, though, even a minor injury would cause the creature to bleed to death.
