’He would sometimes advise me to lay by what was done for several months, and turn my mind to something else, that we might look back at it afterwards with fresh eyes. . . .

’I may mention, because it leads to a general principle of criticism, that, in many cases, the attempt to join truth and fiction did not succeed:  for instance, Mr. Day’s educating Sabrina for his wife suggested the story of Virginia and Clarence Hervey in “Belinda.”  But to avoid representing the real character of Mr. Day, which I did not think it right to draw, I used the incident with fictitious characters, which I made as unlike the real persons as I possibly could.  My father observed to me afterwards that, in this and other instances, the very circumstances that were taken from real life are those that have been objected to as improbable or impossible; for this, as he showed me, there are good and sufficient reasons.  In the first place, anxiety to avoid drawing the characters that were to be blameable or ridiculous from any individuals in real life, led me to apply whatever circumstances were taken from reality to characters quite different from those to whom the facts had occurred; and consequently, when so applied, they were unsuitable and improbable:  besides, as my father remarked the circumstances which in real life fix the attention, because they are out of the common course of events, are for this very reason unfit for the moral purposes, as well as for the dramatic effect of fiction.  The interest we take in hearing an uncommon fact often depends on our belief in its truth.  Introduce it into fiction, and this interest ceases, the reader stops to question the truth or probability of the narrative, the illusion and the dramatic effect are destroyed; and as to the moral, no safe conclusion for conduct can be drawn from any circumstances which have not frequently happened, and which are not likely often to recur.  In proportion as events are extraordinary, they are useless or unsafe as foundations for prudential reasoning.

’Besides all this, there are usually some small concurrent circumstances connected with extraordinary facts, which we like and admit as evidences of the truth, but which the rules of composition and taste forbid the introducing into fiction; so that the writer is reduced to the difficulty either of omitting the evidence on which the belief of reality rests, or of introducing what may be contrary to good taste, incongruous, out of proportion to the rest of the story, delaying its progress or destructive of its unity.  In short, it is dangerous to put a patch of truth into a fiction, for the truth is too strong for the fiction, and on all sides pulls it asunder.’

To live with Edgeworth must have been to enjoy a constant mental stimulus; he could not bear his companions to use words without attaching ideas to them; he did not want talk to consist of a fluent utterance of second-hand thoughts, but always encouraged the expression of genuine opinion.