Who invented supply and demand?

A blog reader emails me:

Hi-

I'm a devoted reader of your blog, and I use your textbook with the Boston Public School seniors whom I teach in Brighton. They love your book. Thanks for doing it!

Today in class, they asked two great questions that I couldn't answer, so I turn to...

Question 1. Who invented the supply and demand graph? (I said Heilbroner. Is that right?) 1B. What is there to read about the history of the graph itself? That is, when was it created? By whom? Was it adopted immediately? etc.

Question 2. Why is quantity on the X axis? Shouldn't it be on the Y axis?

Thanks again for all you do,

[name withheld]
Public School Econ teacher

The model dates long before Heilbroner. Here is the answer according to Wikipedia:

the partial equilibrium supply and demand economic model [was] originally developed by Antoine Augustin Cournot (published in a book in 1838) and thirty years later broadly publicized by Alfred Marshall.

That sounds right to me. I should note that Marshall's 1890 text was the standard for its day. You can find Marshall online, but unfortunately the online version omits the footnotes. In quickly looking through the copy I have on my shelf, I found supply-and-demand graphs in footnotes, but none in the main text. You can, however, read the discussion of supply and demand in Book V.

On the axis question: The instructor is right that, given the way we now teach supply and demand, it makes more sense to have price on the horizontal axis. The price is viewed as the variable that determines quantity supplied and quantity demanded, and we usually put the dependent variable (which here is quantity) on the vertical axis.

So why is it switched? Here is a guess. The early economists may have been imagining that, in the very short run, a given quantity of goods was supplied to the market (an agricultural harvest, for example). The supply curve is then vertical, and the price adjusts to ensure that quantity demanded equals this exogenous quantity supplied. So, in this very short run, the price seems more like the dependent variable. Now, however, the choice of axes is based more on historical convention than logic.

I am not an historian of economic thought, so these answers may be off base. But I am sure the commenters will correct me if I am mistaken.

Update: A comment directs us to a good article on the topic from the Richmond Fed.

Update 2: My Harvard colleague Robert Barro emails me his insights on the matter:

As I recall, Hicks in Value and Capital thought in terms of demand price and supply price. The demand price is how much a person was willing to pay for an additional unit of goods (starting from some initial quantity, Q). The supply price is how much a producer would have to be paid to provide an additional unit of goods. This construction--which I think comes from Marshall--makes it natural to have P on the vertical axis and Q on the horizontal.