Its hard to explain without going into my particular data.  It's on brain
cells. There's two kinds of cells, in many parts of the brain, for several
treatments.  What I'd like, is to have the x-axis be the areas of the brain,
the y be the cell number, and then have the bars each be stacked with the two
cell types.  Only with the several treatments, it's hard. Like i'd like the
control to be maybe black/gray , then one drug be blue/lt blue, then the
other drug be red/pink, and i'd like to have the three groups for each brain
part.

Only I can't get the new data series to stop stacking.