wwh
Although far from an expert on the matter, I don't believe hot spots are localised pods of higher radioactive decay. Let's play visual games...
Hold your hands up level, palms down, in front of you to represent two plates. Bring your fingertips together (convergenet crustal movement), then slide one hand under the other (subduction). The leading edge of the plate that's been subducted (fingertips of your lower hand) starts to melt - note this occurs below and to one side of the impact zone. Because it's hotter (simply through being closer to the fire in effect), the newly melted part is less dense than the surounding upper mantle and crust. Being less dense, it rises towards the earth's surface (ie up towards the middle of the palm of your top hand). Once it hits the earth's surface you've got a hot spot.
The upwelling may never make it to the surface if (frinstance) it is not significantly hotter than the surrounding material or if it cools before it gets there.
A lava lamp is an excellent example of how it all works. The waxy stuff is denser than the surrounding liquid when cool so initially sits at the bottom of the flask. Once heated (and thus less dense), gobs of the stuff rise to the surface where they cool and sink again.
When I last studied the subject, it was thought that the molten part of the interior earth is a function of radioactive decay and gravity. It's a hangover from the earth's molten beginnings. The fact that we're still hot and some other planets aren't is (I think) a product of the earth's size (and thus surface area). Surface area relative to volume increases in inverse proportion when talking of spheres - that's why babies lose so much heat through their heads. Thus the smaller planets have all undergone a greater rate of cooling and the bigger ones are still boiling clouds of gas.
stales
