nary lease."* In this way the actual demesne land would steadily tend to diminish, and at Kingham we find the New College terrier of 1639 summing it up as no more than 160 acres. The manor, in fact, in the older sense of the word, is beginning to break up. Some at least of these free tenants must have had houses and plots of ground, besides their shares in the open field, which were not in existence at the time of the Domesday survey. To these we may perhaps ascribe the completion of the newer street to the south-east, and of the upper end of the older one. Thus the village would have been practically complete, and occupying pretty well the whole of its present site, in the prosperous days of Edward I. And it is a fair conjecture that it remained in much the same condition, and with much the same population, varying no doubt from time to time with healthy and unhealthy periods for several centuries. The number of houses marked on the map of 1828, though not strictly accurate, shows well how many have only been built within the last half-century.

There is one more point to be considered before I close my account. Nothing is said in the Hundred Rolls of an increase of pasture and meadow; yet the extension of area and of population would suggest that there must have been such an increase. Let us see where there was room for it. I have said nothing as yet of a long stretch of meadow lying to the south-east of the village, beween the mill-pond or mill-lake, as it was called,-i.e., the mill-stream,- and the ditch which marks the line of the natural course of the brook before the mill was made. This meadow bears, and always has borne, the strange name of the Yantell, a word of which so far I have failed to discover the origin. This meadow, part of which I have some reason to suppose was originally attached to the Marsh common or waste, and which has no great reputation for the quality of its hay, may have begun to be utilized at that time. I may add, while I am on this ground, that here, just on the village side of the mill-stream, is a small field of an acre or so called in the documents the Conygree, and by our folks still called Kuniger; i.e., the rabbit warren of the demesne. Rabbits as an article of food began to be prized

* Vinogradoff, p. 327.

t † Yantall in the terrier of 1639.

about 1250,* so that we may refer the origin of this to the same period.

But where are we to find room for increase of pasture? Something might be gained by the enlargement of the open field, of which a larger portion would now lie fallow every year; something, too, by the new waste land taken in at the top of the hill beyond the arable. But the increasing community would also need more pasture closes, which could be available all the year round. I think it must have been at this time that a portion at least of the oldest arable of the village was grassed, and turned into permanent pasture. That oldest arable lay, as we saw, to the north-west of the village, and still bears on its surface the marks of ancient ploughing. But my oldest informant remembers it as the common pasture where the cows of all the farmers grazed: it was then separated from the open field by what he called the mound,-i.e., a tun or hedge and ditch sufficiently strong to keep the cattle from straying into the arable. It included, if I understood him rightly, the Marsh (or March) common, the original pasture of the village; but the fact that it was not merged in the Marsh, but known, as we see in the map, by the name of the Back Closes, shows that it was a later addition. The two together would form a wide belt of grass-land extending around the village to north-west, north, and east, and would supply exactly what was needed by a population which must have maintained some hundred plough-oxen, as well as cows and horses. The arable, if I am right, was now pushed away from the village at every point except on the Town Hill, where the land was too excellent to be used for anything but tillage.

The evidence of the Hundred Rolls thus suggests, if it does not prove, that the topography of the village and its land was then in its main features very much the same as it was up to the time of the enclosure in 1843. Between 1279 and 1894 I doubt if it suffered any considerable revolution or felt the economical changes of the period in quite the same degree as some neighboring villages.

W. WARDE FOWLER.

*Rogers, Six Centuries, p. 84. No rabbit warrens are mentioned in Domesday; but they were known to Fleta, temp. Edward I. See Seebohm on Fleta, Village Community, p. 46.

THE CONCEPT OF MARGINAL RENT.

RECENT Contributions to economic discussion indicate a widening acceptance of a formal concept of marginal rent. A distinction between the differential surplus accruing to the owners of superior qualities of land and the monopoly payment secured for the use of marginal land may prove convenient in mere descriptive classification. This does not, however, involve the recognition of an essential difference between marginal and differential rent, or the treatment of incomes received from land on the margin of cultivation as a particular item in distribution subject to a definite law of its own. Such a differentiation seems the result of imperfect analysis, and, in relation to theories of production and distribution, erroneous in theory and harmful in practice. The reasoning of certain of the most careful critics of the Ricardian economics, notably Professor Marshall,* suggests this conclusion with greater or less clearness. In the absence of any explicit statement of the considerations involved, it is proposed in the following paragraphs to describe the concept of marginal rent, to urge that its generic separation from differential rent rests upon the curious neglect of a fundamental element in the law of differential costs, and to suggest the real influence of rent-paying marginal land upon normal cost.† The concept of marginal rent as ordinarily stated assumes the general proposition fundamental to the classical theory of rent, that a given supply of product is derived from the application to land of successive portions, or "doses," of labor and capital of unlike productivity; that the portion last applied is least productive; that the cost of the resultant increment, by the ordinary rule of the market, fixes the cost of the total supply; and that differential rent is measured up from this marginal cost. All rent, it is however asserted, is not

• Principles of Economics (2d ed.), p. 458, et passim. See also "On Rent," in Economic Journal, March, 1893.

↑ "Cost," here as throughout, is employed in the objective sense, better expressed perhaps by Cairnes's "expense."

differential. The dictum, Rent does not enter into normal cost, or, in terms more in harmony with later theories of value, "differential gains do not affect subjective values," is entirely true only in connection with the assumption of the employment of land for a single productive use and the consequent availability of a body of free or no-rent soil. Under modern industrial conditions land is, however, capable of a series of uses, with respect to any but the lowest conceivable of which there is no no-rent land. The poorest or marginal land utilized for any particular purpose is above the margin of utilization or rent-paying land with respect to the next lower purpose. To be retained in the first use, it must yield a "marginal rent" equivalent to that which it would pay if devoted to the second use. This rent enters directly into the cost of the marginal product, and both in method of determination and in relation to normal cost is distinguished from differential rent.

The apparent limitation in the application of the classical law of rent to the classical theory of cost, herein involved, is found repeatedly in economic literature. John Stuart Mill seems to have been the first among English economists to give it expression. In his summary of the theory of value, after asserting that "Rent is not an element in the cost of production of the commodity which yields it," Mill adds,—

But, when land capable of yielding rent in agriculture is applied to some other purpose, the rent which it would have yielded is an element in the cost of production of the commodity which it is employed to produce.†

Jevons cites this passage from Mill with the criticism that "Mill edges in as an exceptional case that which proves to be the rule." He further asks:

But wherefore this distinction between agriculture and other branches of industry? Why does not the same principle apply between two different modes of agricultural employment? If land which has been yielding £2 per acre rent as pasture be ploughed up and used for raising wheat, must not the £2 per acre be debited against the expenses of the production of wheat?

*Patten, Theory of Dynamic Economics, p. 94.

↑ Principles of Political Economy, Book III. chap. vi. (People's ed.), p. 291. Theory of Political Economy (2d ed.), Preface, xlvii, xlviii.

Among recent writers Professor Patten has advanced a similar criticism with greatest force. In his latest work it is described as a "principle ... now accepted in discussions relating to rent," and applied as follows:

If the marginal land used for gardening will yield a rent for wheat, the value of the marginal produce of garden products must equal the cost of the labor employed plus the rent of the land when used for wheat. And if this land is afterwards used for building purposes, the rent which gardeners would pay for the land must be added to the other expenses which the occupiers of these houses must pay.*

Mr. John A. Hobson builds an acute theory of distribution largely upon the premise that:

It is only of unqualified or common agricultural land, in a community which can obtain access to unused land, that it is true that rent forms no part of price. . . . Wherever the worst land in cultivation for a special purpose draws a rent, that rent figures in prices. †

Finally may be noted Professor Wieser's criticism that the overlapping of "secondary" or "derivative" upon "principal" or "original" uses of land constitutes a radical defect in the validity of the Ricardian theory of rent, and that:

Ricardo's proposition that the rent of land does not enter into costs, can be legitimately applied only to land devoted of necessity to one distinct use, such as mines, vineyards, and the like.

The exceptions thus noted to the entire validity of the proposition that rent does not enter into normal cost, while different in approach and in conclusion, obviously rest upon the common recognition of a marginal rent in distribution, distinct from an ordinary differential surplus.

The primary assumptions of the law of rent in its complete form are the variations of different soils in fertility or prox

* Theory of Dynamic Economics, p. 58. See also Stability of Prices, Section II., III., and Premises of Political Economy, p. 22, et passim.

See "The Law of Three Rents," in Quarterly Journal of Economics, April, 1891, pp. 272, 273; also "The Element of Monopoly in Prices," ibid., October, 1891, p. 23.

Natural Value (English translation), p. 209. For similar statements, Andrews, Institutes of Economics, p. 167; Commons, Distribution of Wealth, p. 221; Duke of Argyll, Unseen Foundations of Society, pp. 299, 335.