I found the newspapers of value in the correction of logical assumptions, which frequently appear in American historical and biographical books, especially in those written by men who bore a part in public affairs. By a logical assumption, I mean the statement of a seemingly necessary consequence which apparently ought to follow some well-attested fact or condition. A striking instance of this occurred during the political campaign of 1856, when "bleeding Kansas" was a thrilling catchword used by the Republicans, whose candidate for president was Fremont. In a year and a half seven free-state men had been killed in Kansas by the border ruffians, and these outrages, thoroughly ventilated, made excellent campaign ammunition. But the Democrats had a _tu quoque_ argument which ought to have done much towards eliminating this question from the canvass. On the night of May 24, 1856, five pro-slavery men, living on the Pottawatomie Creek, were deliberately and foully murdered by John Brown and seven of his disciples; and, while this massacre caused profound excitement in Kansas and Missouri, it seems to have had no influence east of the Mississippi River, although the fact was well attested. A Kansas journalist of 1856, writing in 1879, made this logical assumption: "The opposition press both North and South took up the damning tale ... of that midnight butchery on the Pottawatomie.... Whole columns of leaders from week to week, with startling headlines, liberally distributed capitals, and frightful exclamation points, filled all the newspapers." And it was his opinion that, had it not been for this massacre, Fremont would have been elected. But I could not discover that the massacre had any influence on the voters in the pivotal states. I examined, or had examined, the files of the _New York Journal of Commerce_, _New York Herald_, _Philadelphia Pennsylvanian_, _Washington Union_, and _Cleveland Plain Dealer_, all Democratic papers except the _New York Herald_, and I was struck with the fact that substantially no use was made of the massacre as a campaign argument. Yet could anything have been more logical than the assumption that the Democrats would have been equal to their opportunity and spread far and wide such a story? The facts in the case show therefore that cause and effect in actual American history are not always the same as the statesman may conceive them in his cabinet or the historian in his study. In the newspapers of 1850 to 1860 many speeches, and many public, and some private, letters of conspicuous public men are printed; these are valuable material for the history of the decade, and their use is in entire accordance with modern historical canons.