and rhetorically the work is his own. The colonial authorities adopted their own way of composition, and he adopted his. In some sentences we have Bancroft, not Endicott; the nineteenth century, not the seventeenth. Whether the transformation is an improvement or not is not the question; the thing cited is not the original. An accurate historian would no more have issued such a restatement under the shelter of quotation-marks than an accurate theologian would have rewritten the Ten Commandments and read his improved edition from the pulpit. And it is a curious fact that while Mr. Bancroft has amended so much else in his later editions, he has left this passage untouched, and still implies an adherence to the tradition that this is the way to write history. It is also to be noted that the evil is doubled when this practice is combined with the other habit, already mentioned, of relying largely upon manuscript authorities. If an historian garbles, paraphrases, and rearranges when he is dealing with matter accessible to all, how much greater the peril when he is dealing with what is in written documents held under his own lock and key. It is not necessary to allege intentional perversion, but we are, at the very least, absolutely at the mercy of an inaccurate habit of mind. The importance of this point is directly manifested on opening the leaves of Mr. Bancroft’s last and perhaps most valuable book, “The History of the Constitution.” The most important part of this book consists, by concession of all, in the vast mass of selections from the private correspondence of the period: for instance, of M. Otto, the French Ambassador. We do not hesitate to say that, if tried by the standard of Mr. Bancroft’s previous literary methods, this mass of correspondence, though valuable as suggestion, is worthless as authority. Until it has been carefully collated and compared with the originals, we do not know that a paragraph or a sentence of it is left as the author wrote it; the system of paraphrase previously exhibited throws the shadow of doubt over all. No person can safely cite one of these letters in testimony; no person knows whether any particular statement contained in it comes to us in the words of its supposed author or of Mr. Bancroft. It is no answer to say that this loose method was the method of certain Greek historians; if Thucydides composed speeches for his heroes, it was at least known that he prepared them, and there was not the standing falsehood of a quotation-mark.