various scholars for their lords, and later, when this problem had lost
its political importance, the commentators laboured for centuries on the
elaboration of an impeccable system of "ancient emperors"--and to this
day there are sinologists who continue to present these humanized gods
as historical personalities.

In the earlier wars fought between the nobles they were themselves the
actual combatants, accompanied only by their retinue. As the struggles
for power grew in severity, each noble hired such mercenaries as he
could, for instance the landless nobles just mentioned. Very soon it
became the custom to arm peasants and send them to the wars. This
substantially increased the armies. The numbers of soldiers who were
killed in particular battles may have been greatly exaggerated (in a
single battle in 260 B.C., for instance, the number who lost their lives
was put at 450,000, a quite impossible figure); but there must have been
armies of several thousand men, perhaps as many as 10,000. The
population had grown considerably by that time.

The armies of the earlier period consisted mainly of the nobles in their
war chariots; each chariot surrounded by the retinue of the nobleman.
Now came large troops of commoners as infantry as well, drawn from the
peasant population. To these, cavalry were first added in the fifth
century B.C., by the northern state of Chao (in the present Shansi),
following the example of its Turkish and Mongol neighbours. The general
theory among ethnologists is that the horse was first harnessed to a
chariot, and that riding came much later; but it is my opinion that
riders were known earlier, but could not be efficiently employed in war
because the practice had not begun of fighting in disciplined troops of
horsemen, and the art had not been learnt of shooting accurately with
the bow from the back of a galloping horse, especially shooting to the