Thus the first warning we require to take to ourselves is that we must make sure what really is the good we are valuing. In the illustration of last chapter, it was the sack of corn, not the individual grains of corn; it was, that is to say, a group of homogeneous elements considered and valued as a whole. Obviously this is a very different kind of good from, say, a horse or a piano. As durable goods, the latter are, economically, a complex of all the services which they are capable of rendering during their lifetime as goods: their value, therefore, is to be determined by the least use to which their services, one year with another, are put, and not by the least use to which, exceptionally, they are put. Otherwise we should conclude that the utility which a hunting horse may sometimes put forth in drawing a plough, or that which a piano may render at the hands of a schoolgirl, are the marginal utilities determining the value of these goods. Neglect of this consideration led Schäffle to make the objection that, in desert journeys, the traveller’s skin of water, according to our theory, would be measured by the least use to which the water was put: that is to say, the quantity of water employed, say, in washing, would measure the value of the whole skin, while, practically, everybody can see that a good, the possession or non-possession of which meant life or death to the traveller, could not be measured by its washing value. The answer is that here the good which is being valued is the whole water-skin, not the individual drops of water: what measures its value is the amount of wellbeing that would be lost if the _skin_ were lost. If, on the other hand, we were valuing _individual_ cubic inches of water in the skin, or if we were valuing _one_ skin among many, then Schäffle’s calculation would be quite right: that the least use to which the good being valued—the cubic inch or the skin—was put, determined the value of that particular good, the cubic inches or the skin. Similarly, if we ask what is the value of a water-supply to a city, we are putting a different question from “What is the value of the individual gallon of water?” The supply, _as a whole_, is the indispensable condition of a collective human want; the unit of valuation here is not the gallon, but the whole supply. So with the value of a mill stream. We must not confound it with the valuelessness of water as drinking water. What the miller values and pays for is the head of water, and on this the individual cups or gallons used for drinking make no difference. Indeed, we have here one of the exceptional cases mentioned in the beginning of last chapter, the valuation of a single good. The water-supply in the above illustrations cannot usually be put alongside of similar supplies and considered as a member of a stock. Its value is measured by the entire utility which it affords.