And it ties in with the fact that Mill thinks of 'fallacy' as naturally connected to evidence: a fallacy is a case where apparent evidence is not real evidence or apparently conclusive evidence is not actually conclusive, so that a false conclusion is drawn.