in the case of a field where the balance of argument to the farmer in favor of pasture is overwhelming. The position of the margin of transference between different uses may, in other words, be somewhat out of place from the social point of view, and it may be desirable by appeals and propaganda, even conceivably by the devices of State subsidy and compulsion, to push it forwards or backwards in greater or less degree. But it will be necessarily a matter of degree, and nothing could be more foolish than to speak as though there was, or could be, some ideal method of cultivation equally applicable to all lands, without regard to their climatic and other conditions. Needless to say, none of the agricultural experts who sometimes deplore the decline of arable farming are guilty of such foolishness. But the sense of the diversity of nature which is very vivid to them may sometimes be lacking in people who live in towns, and a firm grasp of the marginal notion may serve best to keep the latter from forgetting it. §5. _The Necessity of Rent_. Behind all such detailed applications there lies a more general consideration which deserves attention. The way in which the land of a country is used, the way in which it is apportioned between the countless alternative employments that are possible, is a most important matter, more important perhaps than any questions as to the size of the incomes which particular landowners receive by virtue of their rights of ownership. How is this apportionment effected as things are now? The answer is clear: mainly by the agency of either rent or price. The business which finds it worth while to offer the highest rent or the highest price for any piece of land will, as a rule, be able to command its use. And, with this as the governing principle, an apportionment is secured between shops, offices, factories, agriculture, between the immense variety of different employments covered by each of these broad headings; not a rigid unvarying apportionment, but one which constantly changes as economic circumstances change, and as the margin of transference between different occupations moves hither and thither. This apportionment takes place at present as the result of the independent decisions and bargains of many private individuals, who are thinking mainly of their own interests, and not of those of the community. But this state of affairs might be altered. The land might be nationalized and allocated to its various uses by the co-ordinated labors of a great State department, or some other agency of the collective will. However improbable such a change, it is perfectly conceivable. But what is not conceivable is that any State department should handle the job with a success even approaching that of the present system, unless it continued to use, as its main instrument, the criterion of either rent or price. That a piece of land would yield a higher rent in one occupation than in any other is not conclusive evidence that it