s after half a lifetime, something mysterious happens to his
mental sight. He picks up the book again, and sees a new and profound
significance in every sentence, and he says: "I was perfectly blind
to this book before." Yet he is no cleverer than he used to be. Only
something has happened to him. Let a gold watch be discovered by a
supposititious man who has never heard of watches. He has a sense of
beauty. He admires the watch, and takes pleasure in it. He says:
"This is a beautiful piece of bric-a-brac; I fully appreciate this
delightful trinket." Then imagine his feelings when someone comes
along with the key; imagine the light flooding his brain. Similar
incidents occur in the eventful life of the constant reader. He has no
key, and never suspects that there exists such a thing as a key. That
is what I call a choice absolutely bad.

The choice is relatively bad when, spreading over a number of books,
it pursues no order, and thus results in a muddle of faint impressions
each blurring the rest. Books must be allowed to help one another;
they must be skilfully called in to each other's aid. And that this
may be accomplished some guiding principle is necessary. "And what,"
you demand, "should that guiding principle be?" How do I know? Nobody,
fortunately, can make your principles for you. You have to make them
for yourself. But I will venture upon this general observation: that
in the mental world what counts is not numbers but co-ordination.
As regards facts and ideas, the great mistake made by the average
well-intentioned reader is that he is content with the names of things
instead of occupying himself with the causes of things. He seeks
answers to the question What? instead of to the question Why? He
studies history, and never guesses that all history is caused by the
facts of geography. He is a botanical expert, and can take you to
where the _Sibthorpia europaea_ grows, and never troubles to wonder
what the earth would be without its cloak of plants. He wanders
forth of starlit evenings and will name you with unction all the
constellations from Andromeda to the Scorpion; but if you ask him why
Venus can never be seen at midnight, he will tell you that he has not
bothered with the scientific details. He has not learned that names
are nothing, and the satisfaction of the lust of the eye a trifle
compared to the imaginative vision of which scientific "details" are
the indispensable basis.

Most reading, I am convi