be incorrect, thus correcting it, from various points of view, and
finally, in observing the _*effect_ as it defines itself under this variety
of formulation.

This procedure, adopted in the preparation and judgment of
each new piece of evidence, excludes error as far as our means
conceivably permit. Only one thing more is needful--a narrow and
minute research into that order of succession which is of
indispensable importance in every natural science. ``Of all truths
concerning natural phenomena, those which deal with the order of
succession are for us the most important. Upon a knowledge of
them is grounded every intelligent anticipation of the future'' (J.
S. Mill).[1] The oversight of this doctrine is the largest cause of our
failures. We must, in the determination of evidence, cleave to it.
Whenever the question of influence upon the ``_*effect_'' is raised, the
problem of order is found invariably the most important. Mistakes
and impossibilities are in the main discovered only when the examination
of the order of succession has been undertaken.

In short: We have confined ourselves long enough to the mere
study of our legal canons. We now set out upon an exact consideration
of their material. To do this, obviously demands a retreat to
the starting-point and a beginning we ought to have made long ago;
but natural sciences, on which we model ourselves, have had to do
the identical thing and are now at it openly and honestly. Ancient
medicine looked first of all for the universal panacea and boiled
theriac; contemporary medicine dissects, uses the microscope, and
experiments, recognizes no panacea, accepts barely a few specifics.
Modern medicine has seen the mistake. But we lawyers boil our
theriac even nowadays and regard the most important study, the