dvance. It was a bargain, wasn't it, that I should do your business if
you would help me in mine? Very well; a bargain's a bargain, and we've
both performed our parts. And you mustn't be offended at what I said just
now."

"That I won't! But as to that Raggy Steggles, once those heats are over
to-morrow, I'll--well----"

It was on the following Sunday week that Martin Hewitt, in his rooms in
London, turned over his paper and read, under the head "Padfield Annual
135 Yards Handicap," this announcement: "Final heat: Crockett, first;
Willis, second; Trewby, third; Owen, 0; Howell, 0. A runaway win by nearly
three yards."

III.

THE CASE OF MR. FOGGATT.

Almost the only dogmatism that Martin Hewitt permitted himself in regard
to his professional methods was one on the matter of accumulative
probabilities. Often when I have remarked upon the apparently trivial
nature of the clews by which he allowed himself to be guided--sometimes,
to all seeming, in the very face of all likelihood--he has replied that
two trivialities, pointing in the same direction, became at once, by their
mere agreement, no trivialities at all, but enormously important
considerations. "If I were in search of a man," he would say, "of whom I
knew nothing but that he squinted, bore a birthmark on his right hand, and
limped, and I observed a man who answered to the first peculiarity, so far
the clue would be trivial, because thousands of men squint. Now, if that
man presently moved and exhibited a birthmark on his right hand, the
value of that squint and that mark would increase at once a hundred or
a thousand fold. Apart they are little; together much. The weight of
evidence is not doubled merely; it would be only doubled if half the men
who squinted had right-hand birthmarks; whereas the proportion, if it
could be ascertained, would be, perhaps, more like one in ten thousand.
The two trivialities, pointing in the same direction, become very strong
evidence. And, when the man is seen to walk with a limp, that limp
(another triviality), re-enforcing the others, brings the matter to the
rank of a practical certainty. The Bertillon system of identification--what
is it but a summary of trivialities? Thousands of men are of the same
height, thousands of the same length of foot, thousands of the same girth
of head--thousands correspond in any separate measurement you may name. It
is when the measurements are taken _together_ that you have your man