In case you live in a cave, Harvard's president Larry Summers is in hot water for floating the possibility that there are more men who are innately outstanding in the sciences than women. This runs headlong into the academic consensus that the only difference between men and women is that women are more caring and nurturing than their male counterparts who would destroy the world if left to their own devices. OK, maybe the real academic consensus is that there aren't any differences between men and women that the obvious physical ones and any observed differences are due to societal conditioning.
I commented about my own experiences on women in college level physics (there weren't any when I got my degree) on an interesting post at Tom Maguire's. I'm happy to note that women now account for almost 25% of the bachelor's degrees in physics. As to why women are under 50%, I have to offer my succinct answer: I don't know. It could be that more men are innately talented in that field than women, just as I wouldn't be surprised if women weren't better in some other field of intellectual endeavor. I don't think you can just rule it out because you don't like it. Another alternative, one you probably won't hear from a university president, is that the level of teaching at the undergraduate level in math and science is generally wretched (that was my experience) and women are more likely to go into an area of study with better instruction. Again, the accuracy of the hypothesis can't be proven without proper experimentation. At least, that's something they did teach me in those physics classes back in college.
I surreptitiously took a count in my second-semester Atmospheric Dynamics class this morning (graduate level). I had to count surreptitiously because I sit in the front on the theory that it will help me get good grades. My class has 3 women and 13 men.
Atmospheric Dynamics involves lots of physics and some pretty hairy math.