I wish to make very clear precisely how much I mean by the foregoing argument that circular reasoning is involved in saying that A will take the dodo-bones because B will take them. The same question arises for B, and for the others. The real question is as to the cause for any general practice of the sort. Why should A _suppose_ that B will take them? What could bring about such a system of social relations that a general expectation of this sort could arise? Kemmerer undertakes to give an answer in a hypothetical case by the following ingenious assumption (_Money and Credit Instruments_, p. 11): the money consists of an article which formerly had a high commodity value, which has lately entirely disappeared, but the money continues to circulate, through the influence of custom, and because of the demand for a medium of exchange. In this illustration Kemmerer recognizes the historical fact that money has originated from some commodity which had value because of its significance as a commodity. Historically, a great many different commodities have served, and gold and silver finally emerged victors for reasons which need not just now concern us. These historical facts, coupled with the idea that value is, essentially, "something physical,"[116] or coupled with the notion that value arises only from marginal utility, or from labor, have been accepted by the Commodity or Metallist School as sufficient proof that standard money is only possible when made of some valuable commodity. Professor Laughlin seems to think of the whole thing as depending on the value of gold bullion, and to recognize the money-employment as a factor in affecting the value of money only in so far as it draws gold away from the arts, and so raises its value there by lessening the supply.[117] If money originated in a commodity, how is it possible for the commodity value to be withdrawn, and for money still to retain its value?