can be reached by induction alone. Invention of hypotheses, deduction of consequences, and so on, all have their share in reaching such a conclusion; hence even if an inductive order is adopted in approaching the conclusion, the whole evidence for it can not be set forth in this way. A full demonstration of such a conclusion must necessarily involve some other processes than pure induction. If the presentation appears to be purely inductive, the hearers will have a right to infer that certain important steps have been tacitly passed over; and the speaker may feel sure that if such omissions are detected by any of his hearers, they will form an unfavorable opinion, because of his want of candor or of thoroughness.
The Analytical Method.—This method is characterized by the presentation, at least in outline, of the successive steps that have led the investigator from his original field of observation to the invention of various hypotheses, to the recognition of the most successful hypothesis, and if possible to its establishment as a verified theory, following the plan set forth in the earlier part of this essay. This method is therefore most appropriate in the presentation of complicated problems which demand much theoretical supplement to observation, in the exposition of problems regarding which various unlike opinions have been held by different investigators, and before hearers who are fully able to appreciate rigorous scientific discussion. The essential feature of this method of presentation is that it should preserve the demonstrative quality of the investigation that it represents, and that it should proceed in such an order that the hearers may form a critical opinion as to the value of the conclusion reached at its end. Hence, just as in the usual presentation of a geometrical problem, so in an analytical presentation of a geographical problem, the conclusion or theorem to which the demonstration leads, is advisedly stated not only at the end, but also at the opening of the speaker's address, in order that the hearers may bear it in mind while observed facts, invented hypotheses, deduced consequences, and so on, are all set forth in proper sequence. Only when thus aided by being told the end at the beginning can hearers, who are not familiar with the problem under discussion, really form a competent and critical opinion as to the thoroughness with which it has been investigated.
In view of the short time at a speaker's disposal, the analysis of a complicated investigation can of course be presented only in abstract; but by careful selection of the chief points, it is possible not only to set forth in analytical fashion the leading facts and the most important hypotheses, but also, by impartially confronting the consequences with the facts, to exhibit with convincing clearness the grounds for the final acceptance of one hypothesis and the rejection of its competitors. It should be recognized that while a speaker is thus concerning himself largely with the discussion of past processes, he is for the time being a