putting words in the mouths of historical characters which they never uttered, or did not speak in the given connection.

Eminent examples of transgression are easy to find. Walter Scott brings Queen Elizabeth and Leicester to the front of the stage in Kenilworth. Imaginary conversation is put into their mouths, although the sentiments expressed seem to be consistent. In Woodstock the adventures of Charles II are acted by that royal person, not only with fictive words, but with an entire change of scene from that of the true history. In the more artistic portions of Scott's work, however, the actual personages appear more by description and less in dramatic form. In the Talisman King Richard I constantly plays a talking part, while the history of the third crusade is totally disarranged to suit the purposes of the writer.

So it goes on from Alfred the Great to President Garfield, the authors priding themselves on what the theologians would call the "historicity" of their work. Abraham Lincoln is made to participate in the affairs of a youth otherwise unknown to fame. This proceeds not simply by way of description, but by conversation with the hero and with others in fictitious situations in the usual Lincolnesque dialect. In the effort to make the story seem true the Freeport episode of Lincoln's sensational campaign is related with great circumstantiality, in fact almost identically as found in Miss Tarbell's life of Lincoln. At the same time words are put into Lincoln's mouth which are acknowledged to

be fictitious, but are supposed to be characteristic. Such is the source of all myth.

General Grant furnishes the title and fills a prominent rôle in a recent tale of love and war. His character is painted sympathetically and the descriptive matter is historically unobjectionable, but this great man is made to talk invented words to invented characters for the sake of a love story. It gives one the same feeling as when the liontamer brings out the patriarch of the desert and makes that venerable king of beasts jump through hoops and perform on milking stools. In the "Conqueror" Hamilton is constantly in the foreground. The descriptive portions of the narrative are not only historically careful, but extremely interesting. Certain readers may skip the chapters on the theory of American finance and the founding of the national bank, but the political situation is always vividly described. Why was it necessary, then, to make Hamilton play a talking part, and thus put words in his mouth and create situations which are manifestly fictitious? His biography was already sufficiently romantic.

There is, perhaps, a gift of insight which can recreate the thoughts and words which real historic characters might have said upon imaginary occasions, but too many novelists depend on phrases borrowed from recorded episodes, expecting that these disconnected sentiments will lend reality to an artificial plot. As an inevitable result one class of readers will always associate real personages with untrue incidents, while another class is revolted at

This is

the transparency of the literary dodge. not creative genius, but the art of the stencil plate.

There is a place for the historical novelist, but if he wishes to show true inventive genius let him create, not borrow his characters. Let his fictive personages be so described and so speak that they shall reflect the very image of the age in which they move. Let the passions and ambitions of the human race be depicted in the language and the garb of every century since the world began, and historians and reading world alike will join in the applause. To reach this end true art demands that real historical personages and events shall form the background, not the players and the plot of fiction.1

1 The substance of this chapter was first published in the Book Lover's Magazine, now extinct.

APPENDIX

A SELECTED BIBLIOGRAPHY OF HISTORICAL RESEARCH.

For about two centuries writers of various nations have been steadily discussing the nature of history and the object of its study, with the consequence that the volume of literature is enormous. A part of this material is valuable for the history of changing conceptions of history, while a much larger portion is devoted to the division of history into periods and the suggestion of plans for its teaching and exposition. In the remaining portion of this mass are to be found the works which treat of methods of historical research. Among these the more valuable are among the more recent, and the beginner may profitably select rather than devour the whole.

GENERAL WORKS ON METHOD OF RESEARCH.

Foremost among these is the Lehrbuch der Historischen Methode, by ERNST BERNHEIM, 6th ed., 1908. This is a voluminous treatment of the materials of history, the theory of criticism, the processes of construction, and the philosophy of history. The later editions have given increased space to the latter topic, but their value depends more on the increased references to newer literature. This volume is much indebted to Bernheim for general plan and method of treatment. LANGLOIS AND SEIGNOBOS, Introduction aux études historiques, Paris, 1898. Translation by G. B Berry, 1898 (now out of print). Without attempting to be encyclopedic this provides a stimulating review of the theory and practice of research. CHARLES SEIGNOBOS, La Méthode Historique appliquée aux Sciences Sociales, Paris, 1901. Treats of the methods to be