tion, no mysterious death or mysterious conflagration anywhere in the world of which the causes did not wind back to these personal sources of evil.
3
Worldwide concentration of this kind on a symbolic personality is rare enough to be clearly remarkable, and every author has a weakness for the striking and irrefutable example. The vivisection of war reveals such examples, but it does not make them out of nothing. In a more normal public life, symbolic pictures are no less governant of behavior, but each symbol is far less inclusive because there are so many competing ones. Not only is each symbol charged with less feeling because at most it represents only a part of the population, but even within that part there is infinitely less suppression of individual difference. The symbols of public opinion, in times of moderate security, are subject to check and comparison and argument. They come and go, coalesce and are forgotten, never organizing perfectly the emotion of the whole group. There is, after all, just one human activity left in which whole populations accomplish the union sacrée. It occurs in those middle phases of a war when fear, pugnacity, and hatred have secured complete dominion of the spirit, either to crush every other instinct or to enlist it, and before weariness is felt.
At almost all other times, and even in war when it is deadlocked, a sufficiently greater range of feelings is aroused to establish conflict, choice, hesitation, and compromise. The symbolism of public opinion
