154 
Closely connected with the fallacy of imperfect induction, or 
rather a form of it, is what I may term the fallacy of negatives. 
It is the case where the non-appearance of the instantia con - 
tradidoria is taken as equivalent to its non-existence : whereas, 
as was observed above, we must be sure that if there were any 
instances to the contrary, we should have heard of them or 
discovered them. This error is obviously sufficient to vitiate 
the whole of an induction. It is inductive in its character; 
but there is a fallacy which ( mutatis mutandis ) penetrates into 
the region of deduction, and which I should call by the same 
name, fallacy of negatives. It consists in taking that which 
is not proved as disproved. The overthrowing of one out of 
several arguments in favour of a certain conclusion does not 
prove that conclusion to be false ; it only destroys one syllogism 
in its favour. It may be a very good axiom for practical pur- 
poses that de non apparentibus et de non existentibus eadem est 
ratio ; but it will never do to lay down that everything not 
proved is false. As an instance of the fallacy of negatives, 
I may allude to the sceptical argument against the fact of 
the Resurrection, that it is not mentioned in Roman records. 
So accurately were they kept, it is said, that such an event 
must have been recorded, and discussed at Rome, either as a 
philosophically interesting fact, or as a religious portent re- 
quiring expiation. Now without referring to the explanation 
furnished by Scripture itself, viz., that the soldiers stated that 
they had slept at their post, and allowed the body to be stolen, 
and that the matter was hushed up by means of a large bribe, 
we may rank this argument under the fallacy of negatives. 
How do we know that the matter was not recorded and dis- 
cussed as alleged ? True, we do not find these records, we 
have no account of these discussions ; but are we to infer that 
there were none ? Is it not probable that hundreds of other re- 
markable events were duly reported and made the subject of 
conversation, of which no record remains at this day ? But 
further, it has always struck me that the argument against a 
total deluge, drawn from the state of the extinct volcanoes in 
Auvergne, exhibits this fallacy. I must not be understood 
to be expressing any opinion, — though I have one, — on the 
subject of the total or partial nature of the Flood ; I am only 
discussing the logic of a particular argument. As I under- 
stand the reasoning, it is this : the appearance of certain vol- 
canic craters in Auvergne is such as to show indubitably that 
they have not been covered with water since their last eruption. 
Now there is no record of any eruption having taken place there 
within the memory of the human race, nor any tradition of 
their ever having been active. Consequently the last eruption 
