_Phæd_. But at first sight, you know, variations in the _value_ of labor appear to affect the value of its product: yet you have shown that the effect of such variations is defeated, and rendered nugatory in the end. Now, is it not possible that some such mode of argument may be applied to the case of variations in the _quantity_ of labor? _X_. By no means: the reason why all variations in the _value_ of labor are incapable of transferring themselves to the value of its product is this: that these variations extend to all kinds of labor, and therefore to all commodities alike. Now, that which raises or depresses all things equally leaves their relations to each other undisturbed. In order to disturb the relations of value between A, B, and C, I must raise one at the same time that I do _not_ raise another; depress one, and _not_ depress another; raise or depress them unequally. This is necessarily done by any variations in the _quantity_ of labor. For example, when more or less labor became requisite for the production of hats, that variation could not fail to affect the value of hats, for the variation was confined exclusively to hats, and arose out of some circumstance peculiar to hats; and no more labor was on that account requisite for the production of gloves, or wine, or carriages. Consequently, these and all other articles remaining unaffected, whilst hats required twenty- five per cent more labor, the previous relation between hats and all other commodities was disturbed; that is, a _real_ effect was produced on the value of hats. Whereas, when hats, without requiring a greater quantity of labor, were simply produced by labor at a higher value, this change could not possibly disturb the relation between hats and any other commodities, because they were all equally affected by it. If, by some application of any mechanic or chemical discovery to the process of making candles, the labor of that process were diminished by one third, the value of candles would fall; for the relation of candles to all other articles, in which no such abridgment of labor had been effected, would be immediately altered: two days' labor would now produce the same quantity of candles as three days' labor before the discovery. But if, on the other hand, the wages of three days had simply fallen in value to the wages of two days,--that is, if the laborer received only six shillings for three days, instead of nine shillings,--this could not affect the value of candles; for the fall of wages, extending to all other things whatsoever, would leave the relations between them all undisturbed; everything else which had required nine shillings' worth of labor would now require six shillings' worth; and a pound of candles would exchange for the same quantity of everything as before. Hence, it appears that no cause can possibly affect the value of anything--that is, its exchangeable relation to other things--but an increase or diminution in the quantity