Nor would they be less so in their effect on the general progress of wealth. If commodities and the materials of capital increase faster than the effectual demand for them, profits fall prematurely, and capitalists are ruined without a proportionate benefit to the labouring classes, because an increasing demand for labour cannot go on under such circumstances. If the value of commodities and the materials of capital increase for some time without an increase of their quantity, the labouring classes must soon be supported on the lowest amount of food on which they will consent to keep up their actual number; and the main part of the population would suffer severely without any proportionate benefit to the capitalists; because the value of their capitals, measured by the labour which they can command, would shortly be incapable of further increase. In either of these cases a decided check would be given to the progress of wealth, which progress must necessarily be the greatest, when the joint product of the capitalist and labourer, which the state of the land and the skill with which it is worked enable them to obtain, is so divided between them, that in the progress of cultivation and improvement any unnecessary or premature fall either of profits or corn wages is prevented. But this can only be accomplished by a proper proportion of the supply to the demand, that is, by an accumulation so proportioned to the actual consumption of produce by those who can make an effectual demand for it, as to occasion the greatest permanent annual increase in the value of the materials of capital. The reader of my last work, in which I laid down as my rule, to admit no principles of Political Economy as just which were inconsistent with general experience, will be aware that the conclusions to which I have here shortly adverted, as following necessarily from the constancy of the value of labour, are almost exactly the same as the conclusions of that work. And the reason is, that although at that time I did not think that the labour which a commodity would command could, with propriety, be considered as a _standard_ measure of value, yet I thought it the nearest approximation to a standard of any one object known, and consequently applied it, on almost all occasions, to correct the errors arising from the application of more variable measures. The conclusions, therefore, of my former and present reasonings were likely to be nearly the same, although the premises might now admit of further correction and illustration, and the conclusions might be pronounced with greater precision and certainty.