Sec. 11] EARNINGS AND INCOME 249

be discounted, but, if discounted, it is practically abolished.
Clearly, then, increase of capital is not income in the sense
that it can be discounted in addition to other items of income.
 If it is income at all, it is income in a very peculiar
sense, and nothing but confusion can result from having
to consider two kinds of income so widely divergent that
whereas one is discounted to obtain capital-value, the other
is not.
We have seen that the increase of capital is at the expense
 of income. It is occasioned by and is equal to
the deficiency of realized compared with standard income.
With this in mind, Edwin Cannan’s definition of income
could be stated as realized income plus the deficiency between
 realized and earned income. But this is earned
income, not realized income.

§11

To put the matter in a practical light, let us imagine the
case of three brothers, each of whom inherits the same
fortune, say, $10,000. Let us assume that interest is 5 per
cent. The first brother invests his $10,000 in an annual
annuity of $500 a year forever. The second puts his in
trust to accumulate at 5 per cent for fourteen years, at
which time, having doubled in value, it is to be invested
in a perpetual annuity of $1000 a year. The third, being
of the spendthrift type, buys an annuity of $2000 a year for
(nearly) six years.
According to the theory here advocated, the first has
a perpetual income of $500 a year; the second has no
income for 14 years, and thereafter an income of $1000;
the third has an income of $2000 a year for 6 years and
thereafter none at all. This mode of viewing the matter
also squares with ordinary business reckoning. ;
On the other hand, according to the theory which regards
 increase of capital as income, although the income
from the first would be the same as we have reckoned it,