[Tris]

Ultra strictly you can never proove a theory. Each time observations agree with the theories predictions you are more certain it is correct. As soon as one reliable observation disagrees, you know that it isn't.