The proposition actually proved is, that the event did not 
happen in a particular way — viz., by arrest of motion : a pro- 
position by no means incompatible with the perfect truth of 
the narrative of Joshua. 
These three fallacies appear to me to be those which are 
most commonly to be found in the Deductive logic of Scepti- 
cism. That other violations of the law of Universal Truth, as 
I have called it, occur in sceptical writings and arguments, is 
highly probable, if not morally certain ; you will observe that 
all such false reasoning derives its falsity from the regard- 
ing as the portion of a class placed in a certain relation some 
class or individual apparently but not actually belonging to 
that class. 
I come now to the fallacy of Induction : the neglect of the 
“law of uniformity ” The individual case from which the in- 
duction starts must be, according to this rule, the adequate 
representative of a class ; otherwise there can be no uniformity 
whatever. A false induction, therefore, is made where a re- 
lation between class and class is inferred from the relations of 
an individual not really representing, but only seeming to re- 
present, one of those classes. There is no branch of science, 
I suppose, in which errors of this kind have been more rife, 
than geology. A number of facts having been carefully and 
patiently accumulated, geologists proceeded to their induction, 
and arrived, as they thought, at irrefragable universals, in- 
compatible with the truth of the Scripture narrative. But 
their store of facts was not exhaustive. Some new and un- 
expected discovery has completely modified a proposition once 
regarded as almost axiomatic. I need only refer to the effect 
of the Eozoon Canadense on the appropriateness of geologic 
nomenclature; and the declaration of one of its most eminent 
professors, that the whole science must be remodelled. 
This fallacy was the one against which the old Induction, 
“per simjplicem enumerationem , ubi non reperitur instantia 
covtradictoria /’ failed to guard. It is not the multiplying of 
affirmatives, and the absence of negatives, that constitutes a 
valid induction : it must be made clear also that if any nega- 
tives existed, they would be present ; that the instantia 
contradictoria would be sure to be forthcoming if there were 
one. And thus we find that, to attain truth, we must (as 
Bacon saw) either be able to interrogate nature by arranging 
circumstances for ourselves, and so making an experimentum 
crucis — a hand-post experiment — or resort to some method of 
inquiry which shall eliminate all that is unimportant, and show 
us what is the real representative of the class whose relations 
we may be desirous to investigate. Logicians reduce these 
