Grim linked to an article on new research suggesting that some information acquired in the human lifetime is passed on to offspring, what is called epigenetics. I have been suspicious of epidgenetic claims because
1. It is ill-defined, expanding or contracting at need. I think even among those who make their livings in the research, you will get at least slightly different definitions. It can be broad enough to include prenatal influences - which is not insane because chemical/hormonal exposures in the womb at various developmental stages can make some genes more likely to express, and may even influence what genes are passed on.
2. It is trotted out as a convenient, hand-waving explanation on certain topics whenever we don't like what the genetic research is telling us.
3. Even when it is not full-blown Lysenkoism, it tends to undermine the idea of evolution in general. If there is too much of environment butting in, favorable or unfavorable genes have less pressure. The old sci-fi story "The Lysenko Maze" claimed that we did not see these acquired characteristics being passed on because they are all about unimportant characteristics with little or no survival value. If we make it a significant survival issue for the mice, THEN we will see the traits passed on. Fun idea for a story, but complete nonsense. Survival until breeding (or if you want to stretch a point, survival until your progeny successfully breed) is what was usually being measured, even if not explicitly stated.
4. Some thing else that come to me while I was carmelising onions an hour ago which I can no longer recall. It was a minor, not a killer point, so likely doesn't affect much.
Yet with all that said, I have never denied that a limited amount might still be compatible with plain ol' genetics and might even be likely. In fact, the 150 out of 20,000 (which they think might be low) genes does strike me as low. That it is targeted in particular regions makes sense and has precursors earlier in our phylogeny also seems plausible. WRT skeletons, we have lots of similarity with other creatures. But back to flies? Really? Whoa.
Update: Razib also has an epigenetics in specific circumstances study he links to.