EARTH SCIENCES IN NINETEENTH CENTURY 671 



up either means of investigation. We have learned that both observ- 

 ing and theorizing must be carefully conducted; and we have there-' 

 fore replaced the earlier watchwords, "Go and see," with the later 

 ones, "See and think." We may still give praise to those who apply 

 themselves chiefly to gaining first-hand knowledge of observable 

 facts, but we have learned to give greater praise to those who, on a 

 good foundation of visible facts, employ a well-trained constructive 

 imagination in building ingenious and successful theories which shall 

 bring to sight the invisible facts. We have been longest familiar with 

 the need of theory in those branches of our subject which have, by 

 reason of association with mathematical problems, traditionally 

 employed deductive methods in their discussion, as in earth-measure- 

 ment ; we are least familiar with it in those branches that have until 

 lately followed for the most part inductive or even only empirical 

 methods, as has so generally been the case with geography. 



For example, in the study of the tides, already referred to, how 

 unanimous we are as to the inadequacy of inductive methods; how 

 universally we accept the marvelous theoretical scheme of interaction 

 between planet and satellite, deduced from tidal theory; how we 

 admire its extension to the supposed relation of the inferior planets 

 to the sun ! But in general geography, how little attention has been 

 given to the deductive and systematic consideration of its many 

 problems; how many geographers still look rather askance at those 

 of their number who propose to treat geographical problems through 

 theory as well as through observation! It seems to me clear that, 

 while the earlier progress of geography waswery largely inductive, 

 the later progress has been largely determined by a free acceptance 

 of deductive as well as of inductive methods, and that geography as 

 well as geology is todS,y profiting greatly from the use of our faculty 

 of insight as well as of outsight. 



The objections that are not infrequently urged against the employ- 

 ment of indirect, inferential, as well as of direct, observational, 

 methods in certain branches of our science come from two sides. 

 On one side is a misapprehension as to the nature of our tasks, a 

 belief that our work may really be largely inductive, that observation 

 alone will suffice, if patiently continued, to discover all pertinent facts. 

 This is a serious mistake; there is everywhere more unseen than seen. 



