habitats, that ought never to have been collected at all.* In 

 such lists a single specimen of a rare weed is often given as much 

 space as the commonest native tree, and sometimes even more. 

 At the present time one of the most obvious advantages of 

 studying vegetation rather than flora is that it makes possible 

 much more significant comparisons between different (especially 

 adjacent) regions. By the old floristic method, in order to 

 compare the plant population of two areas it is necessary to 

 determine what species are present in one and absent in the 

 other, which requires pretty thorough exploration ; for the finding 

 of a single specimen of a certain species a few feet inside the 

 boundary of the area from w^hich it was previously supposed to 

 be absent necessitates a readjustment of the statistics. But a 

 determination of what species are more abundant in one area 

 than in another is accomplished more quickly (several of them 

 can be picked out in one day if the areas are close together and 

 well provided with railroads and not more than a few thousand 

 square miles in extent), and is more useful in almost every way. 

 For when the region of greatest abundance of a given species is 

 ascertained, that not only gives the ecologist a clue to its opti- 

 mum environment, but tells the economic botanist where to 

 look for it if it is of any economic importance.! 



* Many if not most collectors do not hesitate to take specimens from fields 

 roadsides, railroad embankments, etc., when the same plants could be obtained 

 just as well in undisturbed habitats near by. The labels ot such specimens, whether 

 they indicate the habitat accurately, or — as is still a common custom, unfortunately 

 — omit it entirely, tell little or nothing about the natural habitat of the species, 

 which is of paramount importance. Worse still, a plant growing in a field may be 

 a little out of its natural range, or larger or smaller or different in some other way 

 from the same species in its native haunts, and in the course of time its descendants 

 may even become specifically distinct, by mutation or otherwise (see Bull. Torrey 

 Club 35: 355-356. 1908); so that placing such specimens in herbaria is likely to 

 mislead the users thereof. Unless one is studying weeds, or the influence of 

 unnatural environments on native species, ruderal plants should be let alone, for 

 cumbering botanical collections with them may easily do more harm than good in 

 the long run. 



t Examples of the floristic method of comparison are common in botanical 

 literature, but the following are among the most recent or easily accessible, or 

 illustrate some special point: MacMillan, Metaspermae of Minn. Valley, 653 et 

 seq. 1892; Beal, Rep. Mich. Acad. Sci. 5: 20. 1904; Harper, Torreya 5: 207- 

 210. 1906; Fernald, Rhodora 9: 158-164. 1907; Gleason & McFarland, Bull. 



