no THE POPULAR SCIENCE MONTHLY
the facts that he comes upon. If he says : " That is a promontory," this simple empirical statement implies that he sees a certain configura- tion of land and water, and recognizes that it possesses the essential features of a typical configuration already known from previous experi- ence, for which the empirical term, promontory, has been adopted by general consent. But if he says : " That is a delta," the statement involves some measure of theory. It implies, as before, that he sees a certain number of features in the land form before him, and that he recognizes their correspondence with the essential features of the con- cept or type, for which the name, delta, has been agreed upon ; but inas- much as a delta is the product of a certain process acting under certain conditions through some unobservable period of past time, the observer has here made a leap into theory, although he may be hardly aware of it. As soon as such a leap is recognized, the visible features of the land form before the observer should be reexamined and stated for the time being in purely empirical terms; that is, in terms based on what is immediately seen, instead of in part on what is inferred. All this calls for fair-minded deliberation, the development of which demands time and training. If this cautious procedure seems slow and cumber- some, it should be practised with respect to various explanatory terms now commonly in use, such as delta, dune, volcano, moraine and so on, until it can be performed with ease and speed. During the progress of such training, a few examples of this elementary kind of analysis should be written out in extenso in the investigator's note book. The number of pages of careful records may at this time serve as a better measure of progress made by a young geographer than the number of miles traversed over hill and valley.
Induction of Generalizations. — When new facts are encountered they are, as has just been shown, more or less consciously compared, in the way of likeness or contrast, with acquisitions of previous experience. As progress is made, groups of similar facts are formed, the several members of a group being alike in respect to certain features that are therefore taken to be essential. An active-minded student quickly gen- eralizes the repeated features by which all the observed members of a group are characterized; thus he conceives an idealized type; and at the same time relegates individual features to a lower rank. As new facts fall into groups already formed and give further warrant for the provisional generalizations previously made, a careful phrasing or formulation of the generalized features should be attempted, with some mention of the way in which individual examples depart from the ideal- ized type. Thus an advancing investigation passes from the recogni- tion of separate facts to the induction of generalized ideas. Certain classes of facts are so fully accessible to observation that the generaliza- tions induced from accumulating records suflBce to provide a reasonable