's trial. I witnessed the scene at
the breakfast table, and, in my opinion, Sir Henry Durwood acted hastily
and wrongly in rushing forward and seizing Penreath. There was nothing
in his behaviour that warranted it. He was a little excited, and nothing
more, and from what I have heard since he had reason to be excited.
Neither at the breakfast table nor in his room subsequently did his
actions strike me as the actions of a man of insane, neurotic, or
violent temperament. He was simply suffering from nerves. It is
important to remember, in recalling the events which led up to this
case, that Penreath was invalided out of the Army suffering from
shell-shock, and that two nights before the scene at the hotel there was
an air raid at Durrington. Shell-shock victims are always prejudicially
affected by air raids.

"Even if the police theory had been correct on this point, it seemed
inconceivable to me that a man affected with homicidal tendencies would
have displayed such cold-blooded caution and cunning in carrying out a
murder for gain, as the murderer at the _Golden Anchor_ did. The Crown
dropped this point at the trial. I merely mention it now in support of
my contention that the case of circumstantial evidence against Penreath
was by no means a strong one, because it originally depended, in part,
on inferred facts which the premises did not warrant.

"Next, the discoveries made in the room where the murder was committed,
and certain other indications found outside, did not fit in with the
police case against Penreath. Superintendent Galloway's reconstruction
of the crime, after he had seen the body and examined the inn premises,
did not account for the existence of all the facts. There were
circumstances and clues which were not consistent with the police theory
of the murder. The probability of the inference that Penreath was the
murderer was not increased by the discoveries we made. I am aware that
absolute proof is not essential to conviction in a case of
circumstantial evidence, but, on the other hand, to ignore facts which
do not accord with a theory is to go to the other extreme, for by so
doing you are in danger of excluding the possibility of any alternative
theory.

"On the other hand, when I sought to account for the crime by any other
hypothesis I found myself puzzled at every turn. The presence of two
persons in the room was the baffling factor. The murderer had entered
through the window in the storm, lig