He very clearly and concisely lays out the case for predictive science:
Darwin realized full well that often we don't have direct evidence, but that doesn't stop us from talking about facts. Indirect evidence can be overwhelming. It can trump direct evidence even! Take a murder, or some other crime against the person. What would lead you to point a finger at a culprit? Sure, eyewitness testimony is going to be very powerful. But we all know that people under strain can be very unreliable about remembering faces. That is not a weakness; it is a very understandable aspect of human nature.
Much better in such cases is the indirect evidence -- the clues, the bits and pieces that point to the culprit. Today, DNA evidence is nigh definitive and with good reason. Any good judge and jury would much sooner know about the molecules on the weapon than the recollections of someone who caught a fleeting glance and whose prejudices may be coloring memory.
It is funny that such an obvious example is lost on people like Ken Ham, who famously asks "Were you there?" Ruse also posits a stunningly simple case for the argument of evolution from homology:
Why -- and this was really powerful evidence -- are the front leg of the horse, the arm of the human, the wing of the bat, the flipper of the porpoise, the paw of the mole, all apparently molded from the same bones connected in the same order when the functions that these forelimbs serve are so very different? Because these animals share common ancestors, and evolution took the early forelimbs and shaped them in different ways, according to need.
Ruse does not share my religious convictions but he writes very easily and clearly. Read the whole article.