It would not merely indicate, as the author states, in which of two commodities varying in relation to each other, at different periods, the variation had taken place;[76] but it would express the precise amount of the variation; that is, if it appeared by documents that the price of a yard of cloth of a certain quality four hundred years ago was twenty shillings, and its price at present was only ten shillings, it would follow, that the estimation in which it was held, or its value, had fallen one-half; because, as all commodities are, by the supposition, produced by labour alone, the sacrifice with which it could be obtained, the necessary conditions of its supply, or the elementary costs of its production, had diminished one-half. The variations of a commodity, in relation to this kind of standard, would further show, with great exactness, the variations in its power of commanding all those commodities which had not altered in the conditions of their supply, or the elementary costs of production. If a commodity rose or fell in this standard price, at different periods, it would necessarily rise or fall exactly in the same proportion in its power of commanding, in exchange, all those commodities which had not altered in the conditions of their supply, or their elementary costs of production. But still, it will be readily acknowledged, that, even granting all that the author has granted hypothetically to Mr. Ricardo, it is not true that such silver would be an _accurate_ measure of the general power of purchasing. Although the circumstance of its invariability, in regard to its producing labour, would give it a prodigious superiority over all other commodities even in this respect, yet, as the producing labour of many commodities may vary in the progress of society, it is quite impossible that the same quantity of any one object can, through successive periods, represent the same general power of purchasing. This is universally allowed; and as it would be clearly desirable to have _one_ rather than _two_ definitions of value, the question is, whether, both on this account, and on account of the universal language and practice of society, for short periods, it would not be decidedly better to confine the term value of a commodity, when used generally, to the estimation in which it is held, determined by the state of the supply compared with the demand, and ordinarily by the elementary costs of production, rather than to its general power of purchasing. There is very nearly an accurate measure of the former; it is universally acknowledged that there cannot be an accurate measure of the latter; and further, it is most important to remark that, in adopting the former, our language would much more nearly coincide with the ordinary language of society in referring to variations of value, than if we adopted the latter.