You seem to think that random mutations are passed on to offspring. Most mutations, as far as I am aware, are somatic mutations. Meaning they are mutations which occur in an organism's non reproductive cells. In order for mutations to be passed on to offspring, they would have to be germline alterations. This is a major sticking point with genetic engineering today, somatic v/s germline alteration. One only affects the organism in question, while the other is passed on to future generations.

How obtuse can you be? Individuals are born with essentially their whole lifetime DNA fixed. There can be changes in the way that individual genes, enzymes, and proteins are *expressed* as a result of environmental feedback, but the basic DNA doesn't change and the offspring have the same DNA barring germline mutation.

However, if germline mutation occurs then the DNA changes. This is usually fatal - tinkering with a watch often is the last time it runs! But if it ISN'T fatal, then a new recipe has been found for successful life. This will usually be only a little different from the parent, due to the way that genes are expressed in context. You then have an experiment in natural selection.

I think in your Somatic vs germline confusion you're conflating the issues of *treating* genetic defects - where the entire soma needs to be treated (or at least the defective organ, in toto), compared to healthy vs unhealthy offspring. To produce a healthy child, only one germ cell needs to be fixed (or selected from a choice of healthy and defective genomes - some test tube babies are cultured for a few cycles so that their genome can be tested for the presence of a harmful inherited defect). These are all the same issues you are speaking about, but you seem to have the whole issue peculiarly confused.

I really wish this idea could be tested in a lab, but many generations would have to be raised in order for it to be properly evaluated. Maybe by breeding a few thousand generations of fruitflies in total darkness, we could see if it had any effect on eye development.

Duh! what do you think generations of Mendel-inspired botanists and zoologists have been doing for the last century and a half? These things are more impressive if you deliberately choose which parents will breed (selective breeding of animal traits in farming and pets, for example). The shorter the generation time of the organism, the faster the process. For example, one study (by Sol speigelman, reported in "Oxygen" by Nick Lane) to find the minimal gene complement required by a virus. If the culture medium contains not just basic nutritional materials but also a large number of the working enzymes and proteins that are needed for the virusto grow, then with time the organism will drop the parts of its DNA that encode for the machinery to make those proteins. The offspring reproduce faster and use less energy because they are not "wasting" effort making things they could just grab from the culture medium. from an initial genome of 4500 base pairs, the virus simplified its genome to a mere 220 base pairs ("Spiegelman's monsters"), yet continued to thrive, so long as the proteins are supplied. The phenomenon of antibiotic resistance is a powerful expression of evolution in action. Also in Lynn Margulis' excellent book "environmental evolution", read about Jeon's work showing how an infectious and generally fatal bacteria can, within a few years become a symbiotic organelle of a strain of amoebae.

BTW: You say that the Ceolocanth of today has different DNA from that of the Carboniferous era, but considering that we have no DNA from specimens from the past, it's hard to come to that conclusion.

I agree that this cannot be proven, for lack of a deep-frozen Carboniferous coelocanth. However, we can be very sure that it works because when we compare non-functional DNA of close and distant relatives, we can see that random mutations accumulate at a fairly steady rate through time. In this way, we can build family trees of evolutionary distance, even between such disparate entities as humans, oak trees, and fungi. See chapter 11 of Andrew H. Knoll "life on a Young Planet"

I recommend you read all three of the books I mentioned above. They are a fascinating compilation of molecular biochemistry and micro- and macro- evolution.