JBR makes a very good point in explaining why textbooks favor the Walrasian way of drawing the axes. Yet I am still more convinced of the alternative explanation, namely, that price adjustment is much easier to deal with (thanks to the auctioneer it may take place in logic, not real, time, so it raises no issues of real dynamics), while quantity adjustment would be a lot trickier as it necessarily involves real time events.
Anyway, is anybody able to check what Samuelson's Principles (first edition) said on the matter? That textbook really set a universal standard.  
Nicola Giocoli
Sent from my BlackBerry® wireless device from WIND