I don't think Marshall reversed the axes.  The first thing I do with demand   
is to show that, for a rational consumer, demand is simply  a (marginal)   
willingness to pay schedule - which of course relates the (dependent)   
willingness to pay to the (independent) quantity.  The next lesson is that   
market demand is simply the horizontal summation of these individual   
curves.     The individual producer's supply curve, likewise, relates the   
(dependent)marginal cost to the (independent) quantity produced (with the   
usual qualifications) ; while market supply is the horizontal aggregation   
of these. Reading the demand curve "vertically'," as it were, may seem less   
natural than reading it 'horizontally" (for any P, how much will consumers   
demand?) but the former is related to the latter as explanans to   
explanandum.  As far as I'm concerned,  Alfred was right !  
  
  
Kevin Quinn