PRESIDENTIAL ADDRESS. 571 
differential significance, upon which all depends, under the same law; and therg- 
fore it cannot be distinctive of land. 
Let us test the truth of these anticipations. We will begin with Ricardo’s 
celebrated law of rent. Granted that land of a certain quality can be had for 
nothing, and is worth cultivating, land of a better quality will bear a rent 
equivalent to its superior yield to the same labour and capital. But this is 
only to say that the superior article fetches the higher price. If we substitute 
‘clothes’ for ‘land’ and say ‘granted that clothes of a certain quality are to 
be had for nothing and are worth wearing, clothes of a better quality will bear 
a price equivalent to the superior satisfaction or service they would yield under 
the same circumstances,’ we shall gain a certain point of vantage from which 
to examine the nature of the Ricardian hypothesis and the validity of the 
conclusions drawn from it; but we shall not have modified it in any way as a 
theory. If it is true theoretically and hypothetically of land it is true 
theoretically and hypothetically of everything; and even if it should turn out 
in any sense to have a superior degree of truth ‘materially’ when applied to 
land, it is only the more fatally misleading and confusing to announce it as 
‘formally ’ proper to it. And again, even to those most inclined to resent or 
at least to challenge this line of argument, it must at once be obvious that the 
Ricardian law of rent takes no account whatever of the distinction between 
inherent properties of the land and what it owes to the capital sunk in it, or 
the incidental advantages rising from the proximity of an industrial popula- 
tion. All these superiorities will tell in the rent demanded exactly on the 
Ricardian principle, so that the difference between economic ‘land’ and capital 
or advantages of situation is completely merged without in any way affecting 
the statement of the supposed ‘law of rent.’ 
A diagram may easily be constructed in which different qualities of land 
may be represented along the axis of X and their supposed relative fertilities to 
a fixed application of labour and capital along the axis of Y. ‘The ‘ marginal’ 
land will occupy the extreme place to the right. A line parallel to the axis 
of X at the height of the ‘marginal’ yield will mark off in the area above it 
the total rent. This is not a functional curve; for the height of Y does not 
depend upon the length of X, but the units are expressly so placed on OX as 
to produce a declining Y. Any confusion on this subject would be equivalent 
to measuring off the altitude of the highest peak of the Himalayas on the axis 
of Y and then, above successive points on the axis of X, measuring the lesser 
heights, in declining order, of successive peaks till we came to Primrose Hill, 
and subsequently reading the diagram as a functional curve showing, say, some 
such relation as that between the height of mountains and their angular distance 
from the equator. There is a difference, then, between a descriptive and a 
functional curve, and what I have just described is a descriptive curve. It is 
applicable to land or to anything else of which typical units can be arranged in 
ascending or descending order of efficiency. 
But the same figure has been used as a functional curve in connection with 
the theory of rent. Take a given fixed area of land of a certain quality and 
consider what would be its yield if it were ‘dosed’ with a certain quantity 
of labour and capital represented by a unit on the axis of X, and represent this 
yield by a rectangle upon this basis. Then add successive increments on the 
axis of X, and after a time you will have to represent the corresponding incre- 
ments in the yield as declining. Go on till a further increment of labour and 
capital would not produce as large an increment in the yield of this land as 
it would if applied to some other piece of land of the same or different quality, 
or if turned to some non-agricultural business. The last increment actually 
applied is the ‘marginal’ increment; the increment it produces in the yield is 
by hypothesis adequate to induce its application, and therefore it measures the 
distributive share of a unit of it in the product. Draw at its altitude a line 
parallel to OX, and the rectangle represents the share of labour and capital, 
whereas the curvilinear surplus or residue represents rent. 
I cannot stay to show that such a curve ought really to be made to pass 
through the origin, for though very essential to sound theory in general, this 
fact does not affect our present investigation. But, on the other hand, it is 
essential to dwell upon several other considerations. In passing I note again 
