ould be a canon in all historical work that the unconscious witness
is the most trustworthy.

I mean by "unconscious" evidence the evidence afforded by one who is not
interested in the type of action which one is attempting to establish.
Suppose, for instance, you wanted to know on exactly what day a Prime
Minister of England left London for Paris upon some important mission. His
biographer who sets out to write an interesting political life and to
insist upon certain motives in him, will say it is the 20th of June,
because Lady So-and-So mentions it in her diary, and because he finds a
letter written by the Prime Minister in Paris on the 21st. Perhaps it is
more important to the picturesqueness of the detail that the journey
should be a hurried one, and without knowing it the biographer is biased
in that direction. There may be twenty documents from the pens of people
concerned with affairs of State which would lead us to _infer_ that he
left London on the 20th, and perhaps only five that would lead us to infer
that he left on an earlier day, and, weighing the position and
responsibility of the witnesses, the biographer will decide for the
twenty.

But if we come across a postcard written from Calais by the Prime
Minister's valet to a fellow servant at home asking for the Prime
Minister's overcoat to be sent on, and if he mentions the weather which we
find to correspond to the date, the 19th, and if further we have the
postmark of the 19th on the postcard, then we can be absolutely certain
that the majority of the fuller accounts were wrong, and that the Prime
Minister crossed not on the 20th but on the 19th, for we have a converging
set of independent witnesses none of whom have any reason to make the
journey seem later than it was, all concerned with trivial duties, and
each unconscious of the effect upon history of their evidence. It would be
extraordinary if the servant had forged a date, and if we suppose him to
have made a mistake, we are corrected by the equally trivial points of the
postmark and the French stamp and the mention of the weather.

So it is with this manuscript record of the King's Kitchen expenses and of
the several halting-places at which they were incurred. Wherever there is
conflict, it must override all other evidence.

The Clerk of the Kitchen, to whom we owe this very valuable testimony, was
one William of Retford. His accounts were kept in a beautifully neat, but
not very legible, fourte