would probably find yourself wondering before long just where the writer got his authority for this or that assertion; and if, in foot-note or appendix, he should seem to satisfy your curiosity by citing some other biography or history, you would be quite justified in feeling that, after all, your inquiry remained unanswered,—for whence did this second writer get his authority? If The question of authority in a book of history you were thus persistent you would probably get hold of the volume referred to and verify, as we say, the statements of fact or opinion attributed to it. When you came upon them you might find it there stated that the point in question is clearly established from certain of Lincoln's own letters or speeches, which are thereupon cited, and perhaps quoted in part. At last you would be satisfied that the thing must very probably be true, for there you would have the words of Lincoln himself upon it; or, on the other hand, you might discover that your first writer had merely adopted an opinion of somebody else which did not have behind it the warrant of any first-hand authority. In either case you might well wonder why, instead of using and referring only to books of other later authors like himself, he did not go directly to Lincoln's own works, get his facts from them, and give authority for his statements at first hand. And if you pushed the matter farther it would very soon occur to you that there are some books on Lincoln and his period which are not carefully written, and therefore not trustworthy, and that your author may very well have used some of these, falling blindly into their errors and at times wholly escaping the correct interpretation of things which could be had, in incontrovertible form, from Lincoln's own pen, or from the testimony of his contemporaries. In other words, you would begin to distrust him because he had failed to go to the "sources" for his materials, or at least for a verification of them.

How, then, shall one proceed in the writing of history in order to make sure of the indispensable quality of accuracy? Clearly, the first thing to be borne in mind is the necessity of getting information through channels which are as direct and immediate as possible. Just as in ascertaining the facts regarding an event of to-day it would be desirable to get the testimony of an eye-witness rather than an account after it had passed from one person to another, suffering more or less distortion at every step, so, in seeking a trustworthy description of the The superiority of direct sources of knowledge battle of Salamis or of the personal habits of Charlemagne, the proper course would be to lay hold first of all of whatever evidence concerning these things has come down from Xerxes's or Charlemagne's day to our own, and to put larger trust in this than in more recent accounts which have been played upon by the imagination of their authors and perhaps rendered wholly misleading by errors consciously or unconsciously injected into them. The writer of history must completely divest himself of the notion that a thing is true simply because he finds it in print. He may, and should, read and consider well what others like himself have written upon his subject, but he should be wary of accepting what he finds in such books without himself going to the materials to which these writers have resorted and ascertaining whether they have been used with patience and discrimination. If his subject is Lincoln, he should, for example, make sure above everything else, of reading exhaustively the letters, speeches, and state papers which have been preserved, in print or in manuscript, from Lincoln's pen. Similarly, he should examine with care all letters and communications of every kind transmitted to Lincoln. Then he should familiarize himself with the writings of the leading men of Lincoln's day, whether in the form of letters, diaries, newspaper and magazine articles, or books. The files, indeed, of all the principal periodicals of the time should be gone through in quest of information or suggestions not to be found in other places. And, of course, the vast mass of public and official records would be invaluable—the journals of the two houses of Congress, the dispatches, orders, and accounts of the great executive departments, the arguments before the courts, with the resulting decisions, and the all but numberless other papers which throw light upon the practical conditions and achievements of the governing powers, national, state, and local. However much one may be able to acquire from the reading of later biographies and histories, he ought not to set about the writing of a new book of the sort unless he is willing to toil patiently through all these first-hand, contemporary materials and get some warrant from them, as being nearest the events themselves, for everything of importance that he proposes to say. This rule is equally applicable and urgent whatever the subject in hand—whether the age of Pericles, the Roman Empire, the Norman conquest of England, the French Revolution, or the administrations of George Washington—though, obviously, the character and amount of the contemporary materials of which one can avail himself varies enormously from people to people and from period to period.

History is unlike many other subjects of study in that our knowledge of it, at best, must come to us almost wholly through indirect means. That is to say, all our information regarding the past, and most of it regarding our own day, has to be obtained, in one form or another, through other people, or the remains that they have left behind them. No one of us can know much about even so recent an event as the Indirect character of all historical knowledge Spanish-American War, except by reading newspapers, magazines and books, talking with men who had part in it, or listening to public addresses concerning it—all indirect means. And, of course, when we go back of the memory of men now living, say to the American Revolution, nobody can lay claim to an iota of knowledge which he has not acquired through indirect channels. In physics or chemistry, if a student desires, he can reproduce in the laboratory practically any phenomenon which he finds described in his books; he need not accept the mere word of his text or of his teacher, but can actually behold the thing with his own eyes. Such experimentation, however, has no place in the study of history, for by no sort of art can a Roman legion or a German comitatus or the battle of Hastings be reproduced before mortal eye.

For our knowledge of history we are therefore obliged to rely absolutely upon human testimony, in one form or another, the value of such testimony depending principally upon the directness with which it comes to us from the men and the times under consideration. If it reaches us with reasonable directness, and represents a well authenticated means of studying the period in question from the writings or other An "historical source" defined traces left by that period, it is properly to be included in the great body of materials which we have come to call historical sources. An historical source may be defined as any product of human activity or existence that can be used as direct evidence in the study of man's past life and institutions. A moment's thought will suggest that there are "sources" of numerous and widely differing kinds. Roughly speaking, at least, they fall into two great groups: (1) those in writing and (2) those in some form other than writing. The first group is by far the larger and more important. Foremost in it stand annals, chronicles, and histories, written from time to time all along the line of human history, on the cuneiform tablets of the Assyrians or the parchment rolls of the mediæval monks, in the polished Latin of a Livy or