<speculation> The hydraulic transmission of a sharp shock does seem to satisfy in the face of the impressive pattern of the shattered bottom of the beaker, and from what I recall a stoichiometric mix of H & O2 has a pretty sharp brisance when ignited. A cherry bomb won't harm a toilet at all unless it's deep enough underwater so that pressure/shock release can't be quickly relieved by raising the water above it. Don't ask me why I know anything about this, I'm sworn to secrecy even though it was a long time ago. <g> So the less gas there is confined in the system to act as a shock absorber, the more brittle or susceptible to shock it would be, and rather small shocks appear to magnify through hydraulic force distribution. The efficient conversion of explosive energy to momentum in the system could account for the destruction of the outer chamber and the high velocity of the shards even beyond the chamber. </speculation>
<wildspeculation> Since the event seemed to develop 'slowly' at first, how about this - a bubble of H & O2 did develop and ignite, but in a burn still slow and/or small enough to be mechanically absorbed by the system. As the fuel was consumed, then - perhaps even synchronously with a returning shock wave from the walls of the vessel from the initial burn - the bubble collapses down against the site of reactions on the screen, causing a tiny supernova. </wildspeculation> - R.
