audiences of 1918. It will be interesting to reread some of the
cheaper and more popular war stories that carried even you, O
judicious reader, off your even balance not five years ago to-day!

We have always known, of course, that a novel can be highly
popular without being truly excellent. Nevertheless, it is a
valuable discipline to specify the reasons. And it is good
discipline also in estimating the intrinsic value of a novel to
eliminate as far as is possible the temporal and the accidental;
and in particular the especial appeal it may have to your own
private craving--for each of us has his soft spot where the pen
can pierce. On the contrary, if the highly speculative business of
guessing the probable circulation of a novel ever becomes yours,
then you must doubly emphasize the importance of these very
qualities; search for them, analyze them out of the narrative,
equate them with the tendencies of the times, the new emotions
stirring, the new interests, new thoughts abroad, and then pick
best sellers in advance.

Yet in eliminating the accidental in the search for real
excellence, it would be disastrous to eliminate all causes of
popularity with it. That would be to assume that the good story
cannot be popular, which is nonsense. The best books are nearly
always popular, if not in a year, certainly in a decade or a
century. Often they spread more slowly than less solid
achievements for the same reason that dear things sell less
rapidly than cheap. The best books cost more to read because they
contain more, and to get much out the reader must always put much
in. Nevertheless, the good novel will always contain one or more
of the elements of popularity in great intensity. I make but one