Evolution isn't quite as cut and dried as a lot of people think. It's a really gradual process based entirely on genetic mutations. The species that survive are the ones with the mutations most advantageous in a particular environment.
I'll make up an example...
Imagine a rodent-like creature (let's call it RodentA) living in a wooded area. These rodents are all attracted to nectar. However, there are other rodents living in the area who are attracted to the same food source as RodentA. Due to a weird genetic mutation, RodentB is born. RodentB has a long snout and can reach the nectar in vase-like flowers. Because this is a virtually untapped resource, RodentB is very successful. RodentB mates with RodentA, and those born with RodentB's characteristics survive. Those born with RodentA's characteristics are less likely to flourish, given the competition for resources that was previously mentioned. So RodentB surpasses RodentA. In a period spanning several generations, RodentB exists, while RodentA is more likely to have become extinct.
I'm not sure if this helps anyone. I just think a lot of the confusion surrounding evolution is that people think of it in overly simplified terms. Creature A doesn't become Creature B. It's a little more complicated than that.
That's a brilliant explanation of microevolution. However it doesn't really apply to macroevolution, which is one species turning into a completely different species.
Personally I cannot accept the theories of macroevolution. One of the biggest problem for me is Michael Behe's theory of irreducible complexity. If a fish managed to evolve lungs, it would drown. The only way to get from a fish to a land animal is to change at the very least the gills into lungs, the single circulation into double circulation and the 2 chambered heart into a 3 or 4 chambered heart all at the same time. It can't happen one small step at a time because the organism simply wouldn't survive. Bearing in mind that mutations don't really add new genetic information (the closest you're likely to get to that is gene duplication followed by point mutations), I find macroevolution rather difficult to believe.