ighed again. Then the window was sealed in and the
experiment started as soon as possible. At the end of the
experiment the window was taken out immediately and the lamp
blown out and then weighed. The amount burned between the time
of weighing the alcohol and the beginning of the experiment was
calculated from the rate of burning before the experiment and
this amount subtracted from the total burned from the time that
the lamp was weighed before being sealed in until the end, when
it was weighed the second time. For the minute which elapsed
between the end of the experiment and the last weighing, the
rate for the length of the experiment itself was used.

During the experiment there were burned 142.7 grams of 92.20
per cent alcohol of a specific gravity of 0.8163.

A tabular summary of results is given below:

+----------------------+--------+-----------+
| | Found. | Required. |
+----------------------+--------+-----------+
| Carbon dioxide gms. | 259.9 | 251.4 |
| Oxygen " | 278.5 | 274.8 |
| Water-vapor " | 165.8 | 165.6 |
| Heat cals. | 829.0 | 834.5 |
+----------------------+--------+-----------+

Thus does the apparatus prove accurate for the determination of all four
factors.

BALANCE FOR WEIGHING SUBJECT.

The loss or gain in body-weight has always been taken as indicating the
nature of body condition, a loss usually indicating that there is a loss
of body substance and a gain the reverse. In experiments in which a
delicate balance between the income and outgo is maintained, as in these
experiments, it is of special interest to compare the losses in weight
as determined by the balance with the calculated metabolism of material
and thus obtain a check on the computation of the whole process of
metabolism. Since the days of Sanctorius the loss of weight of the body
from period to period has been of special interest. The most recent
contribution to these investigations is that of the balance described by
Lombard,[31] in which the body-weight is recorded graphically from
moment to moment with an extraordinarily sensitive balance.

In connection with the experiments here described, however, the weighing
with the balance has a special significance, in that it is possible to
have an indirect determination of the oxygen consumption. As pointed out
by Pettenkofer