together, but not far enough to kill out all varieties and destroy
nature's perpetual tendency to change.' The point of the solution is
not the invention of an imaginary agency, but an assignment of
comparative magnitude to two known agencies.

III.

This advantage is One of the greatest in early civilisation--one of
the facts which give a decisive turn to the battle of nations; but
there are many others. A little perfection in POLITICAL INSTITUTIONS
may do it. Travellers have noticed that among savage tribes those
seemed to answer best in which the monarchical power was most
predominant, and those worst in which the 'rule of many' was in its
vigour. So long as war is the main business of nations, temporary
despotism--despotism during the campaign--is indispensable. Macaulay
justly said that many an army has prospered under a bad commander,
but no army has ever prospered under a 'debating society;' that
many-headed monster is then fatal. Despotism grows in the first
societies, just as democracy grows in more modern societies; it is
the government answering the primary need, and congenial to the
whole spirit of the time. But despotism is unfavourable to the
principle of variability, as all history shows. It tends to keep men
in the customary stage of civilisation; its very fitness for that
age unfits it for the next. It prevents men from passing into the
first age of progress--the VERY slow and VERY gradually improving
age. Some 'standing system' of semi-free discussion is as necessary
to break the thick crust of custom and begin progress as it is in
later ages to carry on progress when begun; probably it is even more
necessary. And in the most progressive races we find it. I have
spoken already of the Jewish prophets, the life of that nation, and