eous. All evidence is divided arbitrarily into
two great classes, direct and circumstantial. I do not here allude to
documentary evidence, which is somewhere between the two, the validity
of the document being necessarily proved by one or the other. This
classification, as I say, is arbitrary, for he would indeed be a wise
man who could tell us exactly where direct evidence ceases to be
direct, or where circumstantial evidence becomes solely
circumstantial. The two are so interdependent, that it is only by
extreme examples that we can dissociate them. All direct evidence must
be sustained by circumstances, whilst all circumstantial evidence is
dependent upon direct facts.

"Let me give you an example of each, that this may be more clear to
your minds. Let us suppose that several boys go to a pool of water to
swim. One of these is seen by his companions to dive into the water,
and he does not arise. His death is reported, and the authorities,
later, drag the pool and find a body. This is called direct evidence.
The boy was seen to drown, you are told, and your judgment concedes
the fact readily. But is the proposition proved, even though you have
these several witnesses to the actual drowning? Let us see. The body
is taken to the morgue, and the keeper there, an expert in such
matters, makes the startling assertion that instead of a few hours, or
let us say a day, the body must have been immersed for several days.
This is circumstantial evidence. The keeper has no positive knowledge
that this particular body has been under water so long. Still he has
seen thousands of bodies, and none has presented such an appearance
after so short an interval. How shall we judge between such
conflicting evidence? On the one side we have direct evidence which is
most positive. On the other we have circumstantial evidence which is
equally so. Is the original hypothesis proven? Does not the
circumstantial evidence raise a doubt? Certainly. Now let us take
another step. The witnesses to the drowning are called again, and view
the body, and now among ten of them, we find one who hesitates in his
identification. At once we find another circumstance wanting in
substantiation of the original claim. Now we see, that all that was
really proved was, that a boy was drowned, and not at all that it was
this particular boy who was found. But is this even proved? How can it
be, in the absence of the drowned body? Now suppose that, at the last
hour, th