16 
MENDENHALL. 
But I am warned that the brief time during which I can 
claim your attention tonight is quite insufficient for any¬ 
thing like a full exposition of the theme which I have se¬ 
lected, and I must, I fear, somewhat abruptly turn about in 
order that I may leave with you in somewhat more definite 
language one or two thoughts which I have attempted to 
develop by illustration and example. 
Recurring to the unfortunate victim of circumstantial evi : 
dence whose experience was related in the beginning, it will 
be admitted that the judge who charged, the jury who con¬ 
victed, the witnesses who told the truth, and the approving 
public were all in error, in that they failed to recognize that 
there was another way of explaining what had happened. It 
does not necessarily follow that the explanation which ex¬ 
plains is the true one. There are many natural phenomena 
which are in entire accord with more than one hypothesis. 
Indeed, there are some things which may be perfectly ac¬ 
counted for on an infinite number of suppositions, but it does 
not follow that all or any one of them must be accepted. 
There is nothing especially novel in this proposition, but I 
submit that to a failure to keep it in sight must be attributed 
a large measure of the uncertainty of the exact sciences, as 
well as much useless and bitter controversy in science, re¬ 
ligion, ethics, and politics. 
As a sort of corollary to this proposition I suggest that 
many reasoning and reasonable people are indifferent to, if 
not ignorant of, the fact that the value of evidence is greatly 
dependent on the way in which it arranges itself. To many 
this may be made a little clearer if I borrow a phrase from 
one of the most exact of modern sciences and speak of evi¬ 
dence as presenting itself in series or in parallel. Without 
pushing the analogy further, the superior strength of the 
latter arrangement will be evident upon reflection. On an¬ 
other occasion, I have referred at some length to the numeri¬ 
cal representation of the value of testimony and to some con¬ 
clusions which are easily reached. As bearing upon the 
