596 



SCIENCE 



[N. S. Vol. XLII. No. 1087 



method in this portion of the subject as is 

 readily attainable in mathematical geog- 

 raphy, and is usually practicable ia physi- 

 ography; but at any rate it is imdesirable 

 to indicate any condition as the controlling 

 one until all other possible influences have 

 been carefully examined and have been 

 shown to have less weight than that one 

 which has been selected. 



Whether the investigation deals with 

 the settlements of man or his movements 

 and means of communication it is impor- 

 tant that in the first instance problems of a 

 manageable size should be undertaken and 

 thoroughly treated, leaving larger areas 

 and wider generalizations until a sufficient 

 stock of thoroughly reliable material which 

 is in the form in which it can properly be 

 used for wider aims is available. 



The relation of geographical conditions 

 to small settlements can be satisfactorily 

 worked out if sufficient trouble is taken 

 and all possible sources of information, 

 both of present date and of periods which 

 have passed away, are utilized. Such 

 studies are of a real value and pave the 

 way to more elaborate studies, but we need 

 more serious study of these simpler cases 

 both to set our facts in order and to pro- 

 vide a methodical classification of the con- 

 ditions which prevail in this part of the 

 subject. Out of such studies there will 

 grow such a series of terms with well-de- 

 fined associations as will give a real pre- 

 cision to the subject which it seems at the 

 present time to lack. 



The same benefit is to be anticipated 

 from detailed work in relation to man's 

 communications and the interchange of 

 commodities in all their varied relations. 

 Generalized and descriptive accounts are 

 readily to be found, and these are for the 

 most part supported by tables of statistics, 

 all of which have their value and present 

 truths of great importance in geography, 



but the spirit of active research which aims 

 at clearing up thoroughly a small portion 

 of the wide field of geographical activities 

 has unequaled opportunities in the some- 

 what shadowy relations between the phe- 

 nomena which we meet in this part of the 

 subject, for focusing the facts better, and 

 obtaining a more exact view of the ques- 

 tions involved. 



Where the geography of states (political 

 geography) is concerned the same need for 

 original investigation as a basis for gener- 

 alizations may be seen. At the present 

 time there is much said about the various 

 boundaries of states, and in general terms 

 the advantages and disadvantages of dif- 

 ferent boundaries under varied conditions 

 can be stated with fair approximation to 

 accuracy. But I do not know of many de- 

 tailed examinations of these boundaries or 

 portions of them where full information of 

 all the factors involved can be found set 

 out in an orderly and authoritative man- 

 ner, thus forming a sure foundation for 

 the generalized description and providing 

 the means of verifying its correctness or 

 revising it where necessary. 



Perhaps there is really more scientific re- 

 search in geography being undertaken by 

 individiials than I have given credit for, 

 but certainly in geogTaphical periodicals, 

 and in the bibliographies which are pub- 

 lished annually, the amount shown is not 

 large; neither is the number of authors as 

 large as might be expected from the im- 

 portance and interest of the subject and 

 from the activity of those centers where 

 geography is seriously taught. There seems 

 to be no reason why individual research on 

 true scientific lines should not be as active 

 in geography as it is in geology, botany, 

 zoology, or any other branch of knowledge ; 

 and, just as in these, the real advance in 

 the subject is dependent on such investiga- 

 tions rather than on travels and explora- 



