Lisa Randall, a Harvard professor of science, attempts to link up the dinosaur killer meteor strike of 65 million years ago with dark matter. It's an interesting read. Dark matter is mysterious, but it's existence is generally accepted. Observation of the rotation of galaxies, shows them rotating so fast that they ought to fly apart. The equations for speed of rotation of a satellite about it's primary go back to Isaac Newton, and are taught in sophomore physics, which makes them well known and universally accepted. Essentially, if a satellite rotates too fast, centrifugal force makes it fly off into outer space and stop being a satellite. If it moves too slowly, the primary's gravity sucks it down and it stops being a satellite and becomes a crater.
The only reasonable answer to the high rotation speed of the galaxies it to assume they contain more matter than you can account for by counting up the stars in the galaxy and estimating their masses. In fact the galaxies come up way short on visible (light emitting) matter, like short by a factor of two or more. So, it's generally accepted that galaxies, including our own Milky Way galaxy, contain a lot of dark matter that does not show up as stars. Just what form this dark matter takes, is unknown at the moment. Lotta people are working on it, and we may have an answer any time now.
Now the author turns to the great dinosaur killer meteor. She wants to show that the Yucatan impact of 65 million years ago is a cyclical event, reoccurring at intervals of 30 million years or so. She cites studies of meteor craters and plots the number and/or size of known craters vs age. These plots give a wavery line on graph paper, and just eyeballing the line doesn't show any apparent periodicity. She goes into a long discussion about just how much periodicity, as opposed to pure random chance, you need to detect it in a graph. Surprise, she never mentioned the standard mathematical method of determining periodicity in any kind of line, the Fourier transform. Apparently she, a Harvard professor, has never heard of Fourier transforms. Well perhaps that's understandable, Fourier transforms are only taught in electrical engineering, no other branch of science has much need for them. Anyhow, without performing the definitive test for periodicity, the author assumes the giant meteor strikes reoccur every 30 million years and then presses on to explain how the Milky Way has a thin disc of dark matter at it's center, and the solar system passes back and forth thru this dark matter disc as it rotates around the galactic core on a 30 million year cycle. Somehow, passage thru the dark matter disk upsets objects in the Kuiper Belt, dragging them out of their nice circular orbits and tossing them down toward the sun in narrow elliptical orbits. Every so often one of them hits the earth, giving us a dino killer event.
It's an interesting read. I also think it's a long stretch.