The latent influence by which the values of things are made to conform in the long run to the cost of production is the variation that would otherwise take place in the supply of the commodity. The supply would be increased if the thing continued to sell above the ratio of its cost of production, and would be diminished if it fell below that ratio. If one dollar covers the expense of making one spade, then when a spade, by virtue of a sudden demand, rises in value to one dollar and ten cents, the manufacturers get an extra profit of ten cents. This could not long remain so, because other capital would enter this industry, and so increase the supply that one spade would sell for only one dollar; then all would receive the average profit. If, owing to a cessation of demand for spades, the price fell to ninety cents, then the manufacturers would lose ten cents on each one made and sold. Thereupon they would cease to do a losing business, capital would be withdrawn, and spades would not be made until the supply was suited to the necessary expense of making them (one dollar). In this way, whenever there is a departure of the value from the normal cost, there is set in motion _ipso facto_ a series of forces which automatically restores the value to that cost. So here again we see the nature of an economic law: the value may not often correspond exactly with cost of production, but there is a _tendency_ in all values to conform to that cost, and this tendency they irresistibly obey. A body possessing weight does not move downward under all circumstances (stones may be thrown upward), but the law of gravitation holds true, nevertheless. There is no need that there should be any actual alteration of supply; and when there is, the alteration, if permanent, is not the cause but the consequence of the alteration in value. If, indeed, the supply _could_ not be increased, no diminution in the cost of production would lower the value; but there is by no means any necessity that it _should_. The mere possibility often suffices; the dealers are aware of what would happen, and their mutual competition makes them anticipate the result by lowering the price. Before the electric light was yet known as a feasible means of lighting (in 1878), the mere rumor of Edison’s invention, before it was made public, and long before it became practicable, caused a serious fall in the price of gas stocks. It is, therefore, strictly correct to say that the value of things which can be increased in quantity at pleasure does not depend (except accidentally, and during the time necessary for production to adjust itself) upon demand and supply; on the contrary, demand and supply depend upon it. There is a demand for a certain quantity of the commodity at its natural or cost value, and to that the supply in the long run endeavors to conform.