258 



THE AMERICAN NATURALIST [Vol. XLII 



in- species, viz., degree of morphologic difference, ab- 

 sence of intermediate forms, and constancy, are entirely 

 inadequate. He sees three critical facts in the develop- 

 ment of a new species or any new form: (1) ancestry, 

 (2) habitat or method of origin, (3) structural changes, 

 very rarely functional ones. None of these can be studied 

 adequately anywhere but on the ground. The first two 

 are purely experimental questions. While the most thor- 

 ough field study is a necessary preparation for them, 

 they can never be decided except by experimental 

 methods in the field. Apart from the study of the origin 

 and structure of vegetation, the fundamental task of the 

 ecologist is the experimental investigation of the origin 

 of new forms, the so-called species. 



III. Ecologic Procedure 

 The questions of what a species is, what are species 

 and what are not, and of the origin of species and of 

 forms must then he decided by experiment. The method 

 of attack is easily determined, but the procedure is super- 

 latively slow and laborious, and the demand upon time 

 and patience unlimited. In spite of his inclination, the 

 ecologist finds it necessary to accept as a working basis 

 the species at present distinguished. His first task is to 

 discover those that seem to give the best promise of re- 

 sults under experiment. The procedure consists of two 

 essential processes: (1) field observations throughout the 

 habitat of the species concerned and (2) experiment. The 

 latter is the crux of the whole question. It consists of 

 three steps: (1) the exact measurement by instruments 

 of the original habitat and the new one, (2) the experi- 

 ment itself and (3) the measurement of results, i. e., the 

 determination of the degree of morphologic and histo- 

 logic difference. The detailed procedure in each of these 

 has been given in "Research Methods in Ecology" 1 and it 

 is unnecessary to repeat it here. It is sufficient to indi- 

 cate that experiment should proceed whenever possible 



1 Research Methods in Ecology. 25, 145. 



