Sprave explains the efficacy of his machine by the large arc-length over which his "bullets" are (uniformly) accelerated. Despite his impressive credentials in the War on Zombies, e.g. by weaponizing pasta, https://www.youtube.com/watch?v=BSjdPufiXec, I doubt his analysis. His drill is so powerful (e.g., firing three balls at once) that it should reach muzzle velocity in a single turn. Couldn't he pick up a factor of √2 or so (doubling the energy) with a (n exponential?) spiral that kept a π/4 ∠, say, between the radial arm and the track? (With consequently nontangential emission.) Note the failure mode in his 2nd video, where the ball breaks out prematurely. --rwg On Thu, Dec 19, 2013 at 5:15 PM, Bill Gosper <billgosper@gmail.com> wrote:

(Which I was if you got my rhumb addendum to the sin 1/x discussion) https://www.youtube.com/watch?v=LH8G7atCFSQ

https://www.youtube.com/watch?annotation_id=annotation_2287679251&feature=iv... Neglecting gravity (not by laying the thing flat!), shouldn't it be possible to eliminate the Archimedean spiral stator by replacing the straight rotor with a very particular curve a la a jai alai basket? Reproducibility might be a problem. --rwg (sin(1/x): What about spiralling into a black hole?)