possess vitality, must deal with actual, living realities,with life in some shape, and most commonly with the life of man. Moreover, in attaining the verbal form which is required, it will naturally possess "style." By this is: not necessarily meant a florid or an obtrusive style. In other words, it does not call for "purple patches."

Science, on the other hand, may be regarded as dealing with certain definite data, by means of systematically reasoned processes, whether deductive or inductive, and as making use of rigid methods of verification, in order to exclude all data which are untrustworthy. It follows from this, that in the work of the scientific historian there is no place for "guess-work" on the one hand, nor for "rhapsodies" on the other. It does not follow from this, however, that "the scientific use of the imagination" is not allowable. It is not merely allowable but even indispensable, provided that it is accompanied by verification, and it is a necessary part of historical science, quite as fully as of physical science, where Mr. Tyndall1 so convincingly advocated it.

If now we inquire as to the materials, the methods, and the aims, of the historian, on the basis of the definitions just given, we may perhaps put the case as follows.

The "scientific historian," so-called, in the use of the materials of his history, will be liberal in the extreme, in extending the scope of the inquiry so as to include not only narratives of wars, of peace, of government, and of the minuter features of every-day life, but he will also be rigid in the extreme in rejecting certain definite data which appear not to have the requisite body of proof in their favor. "Facts",-and nothing else,—will be insisted on, as the appropriate materials for history.2

The "literary historian," on the other hand, will be likely to claim the right to deal not only with facts, but with ideas,

""Fragments of science," (Am. ed.), New York: D. Appleton & Co., 1883, p. 125. "To quote from Ranke:-"Ich will nur sagen wie es eigentlich gewesen ist." Cited by Bury, at p. 18 of his "Inaugural lecture," 1903.

thoughts, fancies, and impressions, maintaining that under some conditions he will find in such a field as this the closest approach to a truthful reproduction of his subject.

The scientific historian will insist on submitting all of his data, and all of his processes, to verification, unhesitatingly casting away whatever does not endure this test. So far as the formal organization of his material is concerned, he will at least aim to present a logical chain of reasoning, even if he does not go so far as to insist on reducing the successive steps in the process to mathematical formulae.

The literary historian, on the other hand, is inclined to attach less importance to formal processes. While he would hesitate to go to the extreme of non-logical methods, he will usually prefer that the "skeleton" of reasoned processes should lie below the surface, rather than on the surface.

The scientific historian urges the necessity of approaching the treatment of any historical incident absolutely free from pre-possession, from pre-judgment, or prejudice, or from pre-conceptions of any kind. He maintains also that the treatment must be absolutely "colorless," so far as concerns the presence, in his own mind, of sympathy, of advocacy, of partisanship, of emotion, or of human feeling generally. In other words, the temper and the treatment, instead of being subjective, must be purely objective.

The literary historian, on the other hand, while admitting that a historian who should, as a matter of fact, be absolutely divested of all human feeling, in approaching a historical subject, would be an interesting phenomenon, maintains that, under existing conditions, this is probably an impossibility. He therefore maintains that a recognition of this fact is safer, in the end, than the assumption of an unrealizable ideal. He maintains also, that in going to

1See the consideration elsewhere in this paper, (p. 385), of this quality, (that of being colorless), as advocated by Ranke.

the extreme of "objective" treatment, one runs the risk of presenting perhaps as distorted a picture, as in going to the extreme of subjective treatment.

The scientific historian conceives of motive only to put it under the ban. He maintains that the only defensible position is that of "history for the sake of history," rather than that of history as a means to an end, however laudable. If the historian may be conceived of as holding opinions, they must be those only which he finds that he can, at the close of his prolonged study of the problem, deduce from the data which have been brought forward. In entering on the study of the problem, however, the shell of no tortoise should be barer of hair than his own mind should be bare of opinions, on either side. It should, in fact, be an absolute blank. He furthermore maintains that, whether or not a history, when complete, is interesting to the reader or not, is no concern of his. His business is with the facts alone. He maintains that to recognize any such motive as that of presenting the facts in an attractive form1 is not only aside from his real province, but is likely to prove a most dangerous and misleading factor in the treatment of the subject. His duty is to get the facts included as a part of the permanent record of history, and then trust to time to bring about their general acceptance, in the light of an extended examination of the subject.

The literary historian, on the other hand, while admitting the danger attaching to pre-conceived ideas, maintains that it is sometimes the obvious duty of a man who has already made up his mind in regard to some occurrence, to set down an orderly narrative of the events connected with it. He also maintains that the writer who fails to present his facts in such verbal form as to carry conviction to his readers falls short of his duty, whether in history, in science, or in literature.

"Dressed up"-to quote from the language already cited above, (p. 351).

The scientific historian is satisfied to toil for months without reaching definite results. He maintains that one of the greatest perils in historical narrative is the confounding of "absolute proof" with what is only "a high degree of probability."

The literary historian, on the other hand, holds that in his general summing up it is perfectly legitimate to cite those data which are only "probable", along with those which are certain,-provided always that this distinction is made perfectly clear to the reader.

Along some such lines of distinction as those above indicated would run the division between the varying points of view of the two schools of historians. And yet, as every student of history knows, a comparative study of individual historians does not reveal a cleavage so simple and so unvarying as that above indicated, but rather an inextricably mixed condition of things. One of the complications is frequently to be noted when the same writer has published both a work of historical narrative proper, and an extended discussion of the ideal "point of view" in

'An effective protest can apparently be made against this position, (the citing of "probable evidence, ") even by those writers who admit in other ways the force of the "literary" point of view, on the ground that it fails to distinguish between the conditions existing in the case of human conduct and those which govern in the framing of a historical narrative.

Bishop Butler, (in his "Analogy of religion, natural and revealed, to the constitution and course of nature," first published in 1736), has convincingly shown that "probable evidence," as distinguished from "demonstrative" evidence, must frequently be accepted in lieu of anything else, in deciding on the steps to be taken, in the practical affairs of life. Long after him, Mr. Gladstone, in his paper on "The law of probable evidence and its relation to conduct," (published under the title of "Probability as the guide of conduct," in the Nineteenth Century, May, 1879, v. 5, p. 908-34, and afterwards reprinted in his "Gleanings of past years, "Am. ed., v. 7, p. 153-99), re-enforced the same view, and included some additional arguments in favor of it. Both of these writers succeed in convincing the candid reader that "probability is the very guide of life." (Gladstone, p. 84.)

But the essential difference between the case of the man who uses "probable evidence," in shaping his course of action, and one who uses it in shaping a historical narrative is, that the former has no option, while the latter has. In other words,

a historian is at perfect liberty not to act on the basis of insufficient evidence, and simply omits all reference to it; and the careful historian will follow this course. The instructive instance cited from Mr. Samuel Rawson Gardiner's experience, (at page 391, below), in which he decided, after long-continued examination of certain papers, that they were "unavailable for historical purposes," (English Historical Review, v. 1, p. 520), is worthy of imitation by all other historians.

writing history. Under these circumstances, it is by no means an unheard of occurrence when such a writer is found strongly emphasizing the need of non-partisan treatment in historical composition, while, at the same time, his own historical work reveals a distinctly partisan point of view. And this helps to show us the futility of any very rigid system of applying the labels, "literary" and "scientific."

It is true that time, place, and condition need to be taken into account, in passing judgment on a historian, particularly in regard to what may be considered the conditions inherently favorable for accuracy. While history, like natural science, has been essayed by both ancient and modern writers, at successive stages of the world's development, one can hardly judge Herodotus, writing in the fifth century before Christ, by exactly the same canons as in the case of James Anthony Froude, writing in the nineteenth century after Christ.

Moreover, the historian's own relation to the event needs to be taken into account. Perhaps the bearing of this principle on the question at issue may best be seen from its operation in the case of biography, which is, after all, a form of history. Imagine, for instance, that a poet and artist such as the late William Morris has died, and that a biography of him is needed. In course of time, a "Life" of William Morris, in two volumes, by John W. Mackail,' makes its appearance. What are the circumstances under which this work has been prepared? This is a question which is very satisfactorily answered, from Mr. Mackail's "Preface" where we read as follows: "When the task of writing the life of Morris was placed in my hands, his family and representatives gave me unreserved access to all the materials in their possession. To them, and more especially to his executors, Mr. F. S. Ellis

'Mackail, John William, Life of William Morris. London: Longmans, Green & Co., 2 v. 1899. Mr. Mackail was appointed Professor of Poetry at Oxford, Feb. 9, 1906.