_Def. 5._ Two articles are never exchanged with each other without a previous estimation being formed of the value of each, by a reference to the wants of mankind and the means of production. This general and most important relation to the means of production, and the labour which represents these means, seems to be quite forgotten by those who imagine that there is no relation implied when the value of a commodity is mentioned without specific reference to some other commodity. M. Say, under the head _Valeur des Choses_, observes, “c’est la quantité d’autres choses évaluables qu’on peut obtenir en échange d’elle.”[97] This is a most vague and uncertain definition, and much less satisfactory than the general power of purchasing. M. Storch says, that “la valeur des choses, c’est leur utilité relative;” but this certainly cannot be said unless we completely change the natural and ordinary meaning either of utility or value. Neither M. Say nor M. Storch has sufficiently distinguished utility, wealth, and value. _Def. 6._ The term creation is not here meant to apply to the creation of matter, but to the creation and production of the objects which have been defined to be wealth. _Defs. 11 and 12._ If wealth be confined to material objects, it must be allowed to be peculiarly convenient and useful, in explaining the causes of the wealth of nations, to have some appropriate term for that species of labour which directly produces wealth; and as the principal founder of the science of political economy has used the terms _productive labour_ in the restricted sense necessary for this special purpose, perhaps few objections would have been made to it, if it had not involved all other kinds of labour, however useful and important, under the apparently disparaging designation of unproductive. This is a consequence, no doubt, to be regretted: yet, when it has been repeatedly stated that the term unproductive, as applied by Adam Smith, in no degree impeaches the utility and importance of such labour, but merely implies that it does not directly produce gross wealth, the mere name ought not to decide against a classification for which it appears from experience that it is very difficult to find a satisfactory substitute. In M. Storch’s “_Considérations sur la Nature du Revenu National_,” he does not appear to me to give a correct view of what Adam Smith means by productive labour.[98] The difficulty of classification above alluded to appears strikingly in this treatise. There is some plausibility in the system, and it is explained with ingenuity and ability; but I think that the adoption of it would destroy all precision in the science of political economy. _Defs. 19 and 20._ I have never been able to understand how the accumulation of capital and the difference between saving and spending can be distinctly explained, if we call all labour equally productive.