species of labour.

Suppose that two hundred thousand men, who are now employed
in producing manufactures that only tend to gratify the vanity of
a few rich people, were to be employed upon some barren and
uncultivated lands, and to produce only half the quantity of food
that they themselves consumed; they would be still more
productive labourers with regard to the state than they were
before, though their labour, so far from affording a rent to a
third person, would but half replace the provisions used in
obtaining the produce. In their former employment they consumed a
certain portion of the food of the country and left in return
some silks and laces. In their latter employment they consumed
the same quantity of food and left in return provision for a
hundred thousand men. There can be little doubt which of the two
legacies would be the most really beneficial to the country, and
it will, I think, be allowed that the wealth which supported the
two hundred thousand men while they were producing silks and
laces would have been more usefully employed in supporting them
while they were producing the additional quantity of food.

A capital employed upon land may be unproductive to the
individual that employs it and yet be highly productive to the
society. A capital employed in trade, on the contrary, may be
highly productive to the individual, and yet be almost totally
unproductive to the society: and this is the reason why I should
call manufacturing labour unproductive, in comparison of that
which is employed in agriculture, and not for the reason given by
the French economists. It is, indeed, almost impossible to see
the great fortunes that are made in trade, and the liberality