^ I don't have much time Jonathon but here's a few quick points.
Mutations happen in DNA (ie genes) in all cells, at all times, including 'germ-line' cells (sperm/egg) which will pass these new traits on. At some level mutations are purely random, at another level they are somewhat non-random in two-ways: (i) some DNA errors are less likely to be picked up by repair machinery and (ii) selection due to differential survivability will choose the good mutations to preferentialy survive.
In practise almost all results in the lab, or known results in nature, produce modified versions of existing genes that are either inactive or still have the same biochemical funciton just with a stronger or weaker 'binding strength'. SO as an explanation for the evolution of the nanomachines in our cells mutaitons is a 'well it wouldn't defy the laws of physics' sort of answer.
PS: the most fascinating example of a mutaiton I ever heard of was of a patient who was being treated for a genetic disease (a near-essential gene was inactivated by a mutaiton sometime in the past generations). Nothing worked and then suddenly the patient started getting better. it turned out that the histoircal mutation back flipped by chance in the important tissues and the patient recovered. If anyone knows the details let me know - I've forgotten. ON the molecular levle the patient had a dud 'gizmo' (eg: an enzyme) - almost perfect but it had a flaw. Random chance flipped that flaw back right again and then the whole gizmo started working again.
[This message has been edited by Tranquility Base, 10-25-2002]
™ Version 4.2