Whenever, as in the case of a battle, very many influences concur to the result, it will frequently happen that that result will baffle our predictions. It will also happen that strange coincidences, such as the frequent recurrence of the same number in a game of chance, will occur. But there are limits to these variations from what we regard as probable. If, in throwing the dice, we uniformly attained the same number, or if in war the army which was most destitute of all military advantages was uniformly victorious, we should readily infer that some special cause was operating to produce the result. We must remember, too, that in every great historical crisis the prevalence of either side will bring with it a long train of consequences, and that we only see one side of the picture. If Hannibal, after his victory at Cannæ, had captured and burnt Rome, the vast series of results that have followed from the ascendancy of the Roman Empire would never have taken place, but the supremacy of a maritime, commercial, and comparatively pacific power would have produced an entirely different series, which would have formed the basis and been the essential condition of all the subsequent progress; a civilisation, the type and character of which it is now impossible to conjecture, would have arisen, and its theologians would probably have regarded the career of Hannibal as one of the most manifest instances of special interposition on record. If we would form sound opinions on these matters, we must take a very wide and impartial survey of the phenomena of history. We must examine whether events have tended in a given direction with a uniformity or a persistence that is not naturally explicable. We must examine not only the facts that corroborate our theory, but also those which oppose it.