[trhway]

>What trhway is saying is that any really big, dense clusters of particles in the tube would probably have roughly the same characteristics as the whole tube they were a part of. The particles rotating around the center of a large, dense cluster of particles in the tube would tend to resemble the rest of the tube statistically, so they would tend to roll in the same direction as the tube itself, meaning the axes of rotation of these big chunks would tend to be parallel with each other and parallel to the centerline of the tube.

yep. Just to add that as those chunks has already clumped close together they are forced to rotate much faster to preserve their share of angular momentum - thus we see rotating quasars on the background of seemingly (in our observation timescale) static filament.