To test evolution, we say: if it's right, we should be able to find things which look like an evolutionary pathway between A and B, in that they will have intermediate forms.
When we've been convinced by this and other evidence that evolution happened, then we can look back at the same fossils and say: A did evolve into B, and these are transitional species.
The first is a successful prediction, the second is interpretation. The fossil evidence for evolution is, of course, the successful prediction.
The additional proof is DNA. Upon it's discovery and when we started genome mapping, scientists began aligning the findings against our current taxonomy table. The results matched the predictions. i.e. species that diverged had common DNA traits that could be linked back to specific common ancestors. And species that were deemed 'close' to each other from an evolutionary standpoint and a specific timescale (i.e. Chimps and Humans) had a high level of overlap in their respective genomes.
™ Version 4.2
