I understand statistical mechanics. I understand that PV=nRT has no lower bound on density. Yet I'm bugged by the thought that the densities in interstellar (or even intergalactic) nebulas are so low that most particles will never experience any collisions at all, thus undermining the premise of statistical mechanics. I understand that some of the particles are charged, but dust clumps of molecules are neutral. I looked up the speed of sound in space (300 km/s), I see pictures on APOD clearly showing shock waves in planetary nebula or supernova remnants, or at the boundaries of the solar system where Voyager is heading. Some shock waves seem driven by radiation pressure, others not. On the other hand galaxies pass through each other all the time with seemingly no effect other than gravity. The other day on APOD they even showed the galaxy Messier 64 and said that it is two concentric counter rotating galaxies coexisting. How the heck can that happen without lots of violent collisions between nebula? The visible evidence seems contradictory. It bugs me; my left brain says the particles collide frequently, my right brain says most particles never collide with anything. Here's my question. Are ordinary statistical mechanics sufficient to describe interstellar and intergalactic clouds of gas and dust, or is there something big I'm missing?