But, secondly, the price at which the whole crop sells is determined by the cost of producing that portion which is produced at greatest cost. It is not, of course, meant by this that the market price of corn always accurately corresponds with the cost of this portion. As was explained on a former occasion,1 when it is said that cost regulates price, what is meant is, that this is the point which the price constantly tends to approach the centre towards which it constantly gravitates. This being premised, it will not be difficult to prove that the price of corn is determined by the cost of producing the most costly portion of the general crop. It is clear that the price must at least be sufficient to cover this cost with the ordinary profit. If it were not, there would be no inducement to farmers to continue the production of this portion: a farmer will not continue permanently to produce corn at a loss. Before he invests his capital in his business, he will consider whether he has a fair prospect of receiving the ordinary returns on it if he has not, he will not invest it. But if the price cannot permanently be less than is sufficient to cover with

a Committee of the House of Commons who piqued themselves on their practical knowledge, that reported that a price of 100s. to 105s. the quarter for wheat was necessary to enable farmers to continue the cultivation of their land; less than this not being a remunerative price;' as if the necessary cost of raising corn were some fixed quantity, independent of the character of the soil on which it is raised, or of the point to which cultivation may be forced upon it. On the other hand, it was reserved for a 'theorist' (Ricardo, in his tract on 'Protection to Agriculture,') to discover that corn may be grown not only in the same country but on the same soil at different costs, and that, therefore, the 'remunerative price' will vary with the state of agriculture.

1 Vide ante, p. 94.

ordinary profits the cost of this portion, it is equally certain it cannot permanently be more than sufficient to do this.

This will appear when we consider the following facts that between the worst and the best lands there are soils of every possible degree of fertility— some on which by dint of high culture corn might be raised, but at such a cost, that it would not replace the capital expended in raising it; others in which, though the returns might replace the capital, they would not yield a profit; others again in which the returns would yield a profit, but less than an average profit; and others still in which the returns will just replace the capital expended with average profits, and no more and when we consider further, that no soil at present in cultivation yields as much corn as it might be made by higher cultivation to yield; that in forcing the soil there is a point at which the returns replace with ordinary profits the capital expended and no more, and beyond which, if cultivation were pushed, though it would lead to an increase of produce, yet this increase would not be sufficient to replace the outlay with the ordinary profit: in a word, that there is a point up to which it is profitable to cultivate, and beyond which it is not profitable to cultivate a fact from which it results that even on the most fertile soil the cost of production may attain any height however great. Now, if these several considerations be borne in mind, it will be seen that the price of corn will not, for any long time, remain at a higher rate than is sufficient to

cover with ordinary profit the cost of that portion of the general crop which is raised at greatest expense; for, were it more than this, the extraordinary profit would at once stimulate cultivation; rich lands would be farmed more highly, and lands of a less fertile quality than before would be brought under tillage, and the process would continue till, either by an increased supply the price was brought down to the cost of production, or through the increasing expense of cultivation, the cost of production rose up to the price.1 It follows, therefore, that as the price of corn cannot remain for any length of time at a lower point than is sufficient to cover the cost with ordinary profits of raising the most costly portion, so neither can it permanently remain at a higher point than is sufficient for this purpose. The extent to which cultivation shall be carried in bringing poor soils under the plough, and in forcing the better qualities-what Dr. Chalmers calls the extreme margin of cultivation' -must be determined by the wants of society; but, wherever that margin may be, whatever in the actual state of agriculture may be the cost of raising the most costly portion of the general crop, this will be the regulator of price-the point which it will constantly tend to approach.

I trust I have now established to your satisfaction the two assumptions on which rest Ricardo's theory of rent. Let me once more repeat them :--of the total quantity of agricultural produce raised in a country, different portions, quality for quality, are

1 Vide ante, p. 94, note.

raised at different costs of production; and secondly, the price at which agricultural produce sells is determined by the cost of producing that portion of the general crop which is raised at greatest expense. From these two assumptions, or, as I may now call them, facts, it results, as I have already shown, that, in the cultivation of agriculture in a country like this a surplus value' arises; while, from the principles of human nature brought into play in the traffic for farms, it follows that this 'surplus value' must go in the form of rent to the proprietor of the soil.

§ 3. The theory of rent just set forth explains the phenomenon of rent in the case of all lands on which agricultural produce is raised at less than the greatest cost at which it can be profitably produced; and this description applies to the great mass of agricultural land in a country like England; but it explains it in this case only. It has accordingly been objected to the theory, first, that it fails when applied to new colonies in which none but the best lands, in point of fertility and situation, are under cultivation; where, therefore, since all the corn is raised at one and the same cost, there could, according to Ricardo's theory, be no surplus value; and, secondly, that it fails to account for the payment of rent in the case of the worst lands under cultivation in every country, on which the whole produce is raised at the maximum of cost, as well as in the case of those lands which are too poor for cultivation, but which nevertheless pay rent.

It cannot be denied that the facts are as the objec

tion states them to be; but, if you have fully seized what I said on a former occasion as to the kind of proof by which economic laws are established or refuted, you will understand that this by no means amounts to an invalidation of the theory. That theory, as I have shown you, rests on facts quite as certain as those which are urged against it, and of far wider reach and more important bearing. What the objection proves is, not that the theory is unfounded, but that, over and above the phenomena which it accounts for, there are others, not perhaps properly described as economic rent,' but of a nature closely allied thereto, for which it does not account. It is a case, in short, and at the utmost, of what in physical science is called 'a residual phenomenon,' and is to be treated in the same way—namely, by looking out for some new cause or principle adequate to explain the residual fact.'

1 On the recurrence of a 'residual phenomenon' in physical investigations it always becomes a question whether the theory, which leaves the fact unexplained, is to be retained, accompanied with the hypothesis of some concurrent cause undetected to which the residual phenomenon may be ascribed, or whether the theory should be wholly rejected. But in economic reasoning no such questions can arise. The grounds of the distinction have been pointed out in the third lecture; they are to be found in the different character of the proof by which ultimate principles in physical and economic science are established. The proof of a physical theory always, in the last resort, comes to this, that, assuming it to be true, it accounts for the phenomena; whence it follows that the occurrence of a 'residual phenomenon ' in physical researches necessarily weakens the proof of the laws which fail to explain it, and, if such exceptions become numerous and important, may lead to the entire rejection of the theory. On the other hand, it is always regarded as the strongest confirmation of the truth of a physical doctrine, when it is found to explain facts which start up unexpectedly in the course of inquiry. (Vide Appendix C.) But the ultimate principles of Political Economy, not being established by evidence of this circumstantial kind but by direct