Equipment for Source Work. It is well to concede at the outset that historical method cannot be taught _successfully_ by a teacher who does not know what it means or who has never applied the method, i. e., done some research work. But perhaps nothing would contribute more to the development of a poorly-trained history teacher than to _oblige him to teach the method; he would be forced to learn something about it_! It is because we have not emphasized the method, because we have not required our candidates for positions as teachers of history to know how to investigate--what would we think of a teacher of chemistry who could not direct the work in the laboratory!--that we have so much absolutely impossible history teaching. The question is, then, can a teacher who knows what historical proof means successfully conduct exercises in historical method in a high school? I think there can be no doubt of it. It is being done. To conduct the work successfully a source book, differing in some respects from the majority of source books, is needed. There are two kinds of historical facts: one class can be established by a single source, the other--and this is the more difficult, but at the same time the more valuable as training--can be proved to be true only by the agreement of independent sources or witnesses. For this last kind of work more than two sources treating of the same event are necessary. As the most of the source books are only intended to supply collateral reading, they contain little material that could be used for critical exercises. My source book on Greek history contains some such exercises, and it would be a matter of no great difficulty to supplement the sources in any of the books by two or three extracts dealing with the same topic. Sources in the Class Room. Two exercises a week would be enough for intensive critical work. The sources should, of course, be in the hands of the pupils and the attention of the class should never be allowed to stray from the evidence in the text. It is not necessary that the work should be systematic at the outset or that it should be forced. It might be introduced in a very simple and natural way by an attempt to settle the truth of some point upon which two school texts disagree. It is a common practice, in schools where several narratives are used, to assign different texts to different pupils and in the recitation hour, to compare the statements of the writers. Suppose they disagree? I once asked a teacher who employs this method what she did in such a case. She answered that they discussed the matter, and, if they could reach no agreement as to which statement was correct, they dropped it. A more pernicious practice could hardly be imagined. The class was run into a blind alley and left there! The escape was easy enough, if the teacher had been master of the situation. It offered an excellent point of departure for the introduction of the study of historical method.