ot "a beloved," he was a
true vagabond, and I first met him in the Champs Elysees, just as in "The
Pigeon" he describes his meeting with Wellwyn. Though drawn very much
from life, he did not in the end turn out very like the Ferrand of real
life--the, figures of fiction soon diverge from their prototypes.

The first draft of "The Island Pharisees" was buried in a drawer; when
retrieved the other day, after nineteen years, it disclosed a picaresque
string of anecdotes told by Ferrand in the first person. These two-thirds
of a book were laid to rest by Edward Garnett's dictum that its author
was not sufficiently within Ferrand's skin; and, struggling heavily with
laziness and pride, he started afresh in the skin of Shelton. Three
times be wrote that novel, and then it was long in finding the eye of
Sydney Pawling, who accepted it for Heinemann's in 1904. That was a
period of ferment and transition with me, a kind of long awakening to the
home truths of social existence and national character. The liquor
bubbled too furiously for clear bottling. And the book, after all,
became but an introduction to all those following novels which
depict--somewhat satirically--the various sections of English "Society"
with a more or less capital "S."

Looking back on the long-stretched-out body of one's work, it is
interesting to mark the endless duel fought within a man between the
emotional and critical sides of his nature, first one, then the other,
getting the upper hand, and too seldom fusing till the result has the
mellowness of full achievement. One can even tell the nature of one's
readers, by their preference for the work which reveals more of this side
than of that. My early work was certainly more emotional than critical.
But from 1901 came nine years when the critical was, in the main, holding
sway. From 1910 to 1918 the emotional again struggled for the upper
hand; and from that time on there seems to have been something of a "dead
beat." So the conflict goes, by what mysterious tides promoted, I know
not.

An author must ever wish to discover a hapless member of the Public who,
never yet having read a word of his writing, would submit to the ordeal
of reading him right through from beginning to end. Probably the effect
could only be judged through an autopsy, but in the remote case of
survival, it would interest one so profoundly to see the differences, if
any, produced in that reader's character or outlook over