Theorists who have confined themselves to the examination of exchange value, or, what comes to the same thing, of price, may have succeeded in discovering certain empirical laws of changes in amounts of value, but they could never unfold the real nature of value, and discover its true measure. As regards these questions, so long as examination was confined to exchange value, it was impossible to get beyond the formula that value lies in the relation of exchange;--that everything is so much more valuable the more of other things it can be exchanged for.... Absolutely and by itself, value was not to be understood. It is significant of this conception to state that one thing cannot be an object of value in itself; that a second must be present before the first can be valued. Theory has only very gradually shaken itself free from this misconception, this circle. Where an absolute theory was attempted--such as the labour theory, or that which explained value as usefulness--some logical leap generally reconnected it with the relative conception. Now the validity of this reasoning might be admitted, in so far as it applies to "Crusoe economics"--though Professor Seligman, with strict consistency, insists that even there value arises from a comparison in Crusoe's mind of apples with nuts[26]--by those who would object to its application to value in society. Value there, it would be insisted, is determined through exchange, and does not have any meaning except as a ratio between physical commodities.[27] But even here, it seems to me, the same reasoning must hold. We really do not find a ratio between physical commodities at all. Four gallons of milk exchange for one dollar, or 23.22 grains of gold. The exchange ratio is four to one. But milk is in units of liquid measure; gold in incommensurable units of Troy weight. The ratio, 4:1, is not on the basis of any physical commensurability. If any physical basis of comparison be taken, whether weight, or bulk, or length, or more subtle and less easily measurable physical qualities, the ratio would be found very different. But 4:1 _is_ the market ratio. Now a quantitative ratio is between commensurable quantities. Gold and milk must be, then, commensurable quantities, _i.e._ must have a common _quality_, present in each in definite quantitative degree, before comparison is possible, or a ratio can emerge. This quality is _value_. The difficulty, from the standpoint of logic, is only covered up, and not avoided, if we say with Professor Davenport,[28] "Value is a ratio of exchange between two goods, _quantitatively specified_." [Italics mine.] For the quantitative specification depends on the extent to which the homogeneous quality is present in each of the goods, or, if we assume that the quantitative specification is made before the question of exchange ratio is raised, then