is simply in the adaptation of parts to act harmoniously
for definite ends. Hence if we are allowed these three factors, we can
say that nature _does possess forces adequate to the manufacture of
machines_. These forces are not chemical forces, and the construction of
the machine has thus been brought about by forces entirely different
from those which produced the chemical molecule.

But we have plainly not reached the bottom of the matter in our attempt
to explain the machinery of living things. We have based the whole
process upon three factors. Reproduction, variation, and heredity are
the properties of all living matter; but they are not, like gravity and
chemism, universal forces of nature. They occur in living organisms
only. Why should they occur in living organisms, and here alone? These
three properties are perhaps the most marvellous properties of nature;
and surely we have not finished our task if we have based the whole
process of machine building upon these mysterious phenomena, leaving
them unintelligible. We must therefore now ask whether we can proceed
any farther and find any explanation of these fundamental powers of the
living machine.

It must be confessed that here we are at present forced to stop. We can
proceed no further with any certainty, or even probability. We may say
that variation and heredity are only phases of reproduction, and
reproduction is a property of the living cell. We may say that this
power of reproduction is dependent upon the power of assimilation and
growth, for cell division is a result of cell growth. We may further say
that growth and assimilation are chemical processes resulting from the
oxidation of food, and that thus all of these processes are to be
reduced to chemical forces. In this way we may seem to have a chemical
foundation for life phenomena. But clearly this is far from
satisfactory. In the first place, it utterly fails to explain why the
living cell has these properties, while no other body possesses them,
nor why they are possessed by living protoplasms _alone_, ceasing
instantly with death. Indeed it does not tell us what death can be.
Secondly, it utterly fails to explain the marvels of cell division with
resulting hereditary transmission. For all this we must fall back upon
the structure of protoplasm, and say that the cell machinery is so
adjusted that the machine, when acting as a whole, is capable of
transforming the energy of chemical composition in