Again, what you say here is right, but in my mind, the logic is backwards. If you look at the example you just gave, the cardinal was more brightly colored and that was selected for, right? If that's the case, then the mutation happened first, and selection happened because of the mutation, not the other way around. Saying that mutation follows selection doesn't follow logic for two reasons that I see. 1) Evolution (or selection) doesn't look forward in time. The genes don't say "well I think he would be able to reproduce more successfully if he were brighter, so I think I'll mutate now." The gene mutates, and if it helps him reproduce more, then that trait will be selected for. 2) A mutated gene is one that is different from the genes passed on by either parent (not just phenotypically, otherwise recessive alleles would be lumped into that definition too and we know that it doesn't work that way). So if the F1 generation individual has a mutated gene that neither of the parents have, and he passes that gene onto the F2 generation, it's no longer a mutation because it came from one of the parents. So with the cardinal example, it doesn't change the identification as a species, just like you said. But if the mutation was such that the mutated-gene cardinals had a feeding or reproductive advantage because of their mutation, then after several generations, they MIGHT form a new species. Think about it, if your goal was to have successful offspring and you knew you could, let's say, find food better than other cardinals because of your mutation, wouldn't you seek out other cardinals with the same mutation (so it's more likely you'll pass the advantage on)? Obviously this wouldn't work right away with only one bird having the mutation, but it would have to take place after several generations so that more and more had the mutation.