Januaky 29, 1904.] 



SCIENCE. 



185 



geographical study as forms of glacial dep- 

 osition are already given. Neither the 

 thing studied, nor the agent by which it 

 was produced, nor the method by which 

 the agent is shown to be accountable for 

 the thing, suffices to show whether the thing 

 is of a geological or a geographical nature. 

 This question will be decided, as has 

 already been shown, by the relations into 

 which the thing enters. It would be as 

 unreasonable to omit all reference to gla- 

 cial erosion in a geographical description 

 of Norway as to omit all reference to sub- 

 aerial erosion in a geographical account of 

 our Atlantic coastal plain. 



Nowhere is the cultivation of systematic 

 geographj^ more helpful than in the study 

 of local or regional geography. The truth 

 of this may be appreciated by considering 

 the case of botany. No botanist would at- 

 tempt to describe the flora of one of our 

 states until he had obtained a good Imowl- 

 edge of systematic botany in general. 

 Such knowledge would help him at every 

 turn in his study of a local flora, not only 

 in describing the plants that he might find, 

 and in arranging the descriptions in a serv- 

 iceable order, but also in finding the 

 plants themselves. I believe that a closely 

 equivalent statement might be made with 

 regard to the geography of a state ; and 

 yet there is not, to my knowledge, a single 

 work on regional geography in which a 

 recognized scheme of systematic geography 

 has been avowedly followed as a guide for 

 the treatment of local features. The adop- 

 tion of si^eh a guide would lead to various 

 advantages; on announcing that a certain 

 scheme of systematic geography has been 

 chosen as a standard, the writer of a re- 

 gional work thereby gives notice in the 

 simplest manner to the reader as to the 

 ■ kind and amount of knowledge necessary 

 to understand the work in hand; descrip- 

 tions are made at once briefer and more 

 intelligible than by phrasing them in terms 



of a scheme that is elsewhere stated in full ; 

 relative completeness of treatment is as- 

 sured, for with a systematic list of all 

 kinds of geographical relations at hand, 

 the writer is not likely to overlook any 

 element of the subject that occurs within 

 his chosen region; the reader can easily 

 find any desired topic, not only by means 

 of the table of contents and index, but also 

 by means of the standard scheme of classi- 

 fication in accordance with which all ele- 

 ments are arranged; and finally, books on 

 different regions will come to exhibit a 

 desirable uniformity of treatment when 

 they are based on a common scheme of sys- 

 tematic geography. Although no books of 

 this kind now exist, I do not think it over- 

 venturesome to say that some such books 

 will soon exist, and that they Avill form 

 very serviceable contributions to the litera- 

 ture of our subject. 



The various recommendations that I 

 have made are likely to remain in the air, 

 or at most to secure response only from 

 isolated individual students, unless those 

 who believe that the adoption of these rec- 

 ommendations would promote the scien- 

 tific study of geography are willing to give 

 something of their time and thought 

 toward organizing a society of geograph- 

 ical experts — an American geographers' 

 union. From such a union I am sure that 

 geography would gain strength, but it is 

 not yet at all clear in my mind that any 

 significant number of persons would care 

 to accept the strict conditions or organiza- 

 tion which appears to me essential for the 

 success of such an enterprise. .The most 

 important of the conditions are as follows: 



1. The adoption of some definition for 

 ■ geography that shall sufficiently indicate 



the boundaries as well as the content of 

 this broad subject. 



2. The limitation of membership to per- 

 sons with whom geography as thvis defined 

 is a first or at least a second interest, and 



