A few further musings on the biogeography chapters:
1. After today's discussion, I now want to go and read Mayr (which I've never actually done, embarrassingly), and decide for myself just how firm are the empirical (as opposed to theoretical) foundations of the 'conventional wisdom' that allopatric speciation is the rule.
2. Darwin's own theory of speciation is (in)famously hard to decipher. As was correctly pointed out in class today, much of the biogeography material sounds like it was written by a man who believes in allopatric speciation as the dominant mode of speciation. But the earlier chapters don't read that way. Darwin clearly understood that what we would now call gene flow can prevent lineages from splitting. But his theory of speciation also relies a lot on the 'principle of divergence of character' (see chapter IV). Briefly, this is the idea that the 'fittest' types of organism will be those that produce the most divergent offspring, so that collectively those offspring can occupy a greater number of niches in the economy of nature, thereby allowing more of the offspring to survive than if they were all suited only for a single niche. This principle has several logical and empirical problems with it, although something like it is the basis for one hypothesis to explain the evolution of sexual reproduction (which produces more variable offspring than asexual reproduction).
I wonder if Darwin was drawn to the principle of divergence of character in part because he was afraid that otherwise his theory would predict a world populated only by a very few 'superspecies' rather than the polyglot diversity that we see. There are other ideas that Darwin might have hit on (or emphasized more strongly) to explain why we don't just have a few superspecies, such as the fact that selection pressures vary geographically. But taking that view seriously would've required him to find another way to explain why some species have larger geographical ranges than others besides the simple claim that the wide-ranging species are the 'fittest'. In general, Darwin isn't as sensitive as I expected to the notion that fitness is typically very context-dependent. He talks about the fittest type replacing all other types in its own lineage, and then spreading widely to replace other similar lineages, as if any one type could possibly be the 'best' in very different contexts. Then, to ensure continued maintenance and production of diversity, he appeals to the principle of divergence of character, which introduces an element of self-contradiction. Your offspring can't fill the gaps in the economy of nature (thereby avoiding competition with other species), and at the same time outcompete other species for already-occupied places in that same economy. Your offspring either compete with other species, or they don't.
But Darwin had such a head full of ideas, it's hard to fault him too much if those ideas didn't all quite fit together.
3. A rare instance of Darwin just getting it wrong: Darwin refers at multiple points to species failing to evolve in response to climate change because they all shift their ranges as a group, so that each continues to live with its coevolved fellows. There are a number of problems with this view, starting with the fact that it's false. Empirical research has now shown that, while species do generally migrate north (and up mountain slopes) in warming periods, and the reverse in cooling periods, they do so at very different rates. Many ancient ecosystems therefore were comprised of combinations of species that don't currently exist. But in fairness, Darwin's suggestion is a perfectly reasonable one given the evidence available to him at the time.