Precisely why I do not like the Newtonian corpuscle idea. In fact, I don't believe that there is anything but empty space in between. I like to think that certain quanta are bound on all sides by a process (I like to think of things in terms of computers). I imagine a rule that the process will bind a quantum on all sides, like a ray drawn from infinity that comes from all points in the background. If there is a quantum that interrupts a ray, a quantum that is in its shadow is not bound on all sides and it moves toward the interrupting quantum. This only works if no three quanta are in a straight line.

By the way, this idea of a binding process makes me wonder about a possible relationship to spin and frequency. I mean, it would determine which direction it is being bound from at the moment, and how often it is bound. (Relative to other quanta)