If it's given that new species can arise, then we can draw the "family tree" of the new species. And it's logical to ask, how far back can we go? How many species can we fit under one umbrella?
For example, all mammals have a bunch of things in common. A reasonable explanation is that all mammal species are cousins. But can we convincingly draw a family tree for all mammals? Can we find a place to connect strange half-mammals like the platypus? Here is part of the family tree from the Tree of Life project:
=============== Triconodonts (extinct) | | ============ Monotremata (platypus,echidna) | | | |=========== Multituberculata (extinct) mammalia | | ===| ====== Marsupialia (kangaroos etc) | | ======| === Palaeoryctoids (extinct) ===| === eutheria (whales,mice,elephants)
The Theory of Common Descent is the idea that we can take these diagrams all the way back, until there is one single ancestral species. Every living thing (and every extinct one) would have to fit naturally on just one diagram.
There are three scientific questions here: