As you probably know, the "speed of sound" in an arbitrary medium is just the speed at which disturbances in pressure propagate. For this to be well-defined, the particles have to collide with each other often enough for there to be collective disturbances (rather than just individual fast moving particles). I would have thought that there was so much distance between particles in interstellar space that they never collided, but apparently there are often enough collisions to define a notion of sound. Here's wikipedia:

> The shock arises because solar wind particles are emitted from stars at about 400 km/s, while the speed of sound (in the interstellar medium) is about 100 km/s. (The exact speed depends on the density, which fluctuates considerably.) The interstellar medium, although very low in density, nonetheless has a constant pressure associated with it; the pressure from the solar wind decreases with the square of the distance from the star. As one moves far enough away from the star, the pressure from the interstellar medium becomes sufficient to slow the solar wind down to below its speed of sound; this causes a shock wave.

Just curious why the speed of sound in interstellar medium depends on density, while in the Earth atmosphere, it does not (it is nearly same as on surface, around 10% smaller, on 100,000ft where there is hardly 2% of surface air pressure).

When you can hear it, duh.