GEOGRAPHY AND VEGETATION OF NORTHERN TEORIUA. 175 



ly how nikiiy ut each class. In describing the agricultural output 

 of Florida we do not merely say that corn, cotton, oranges, vege- 

 tables, etc., are raised in the state, but we give the quantity of each 

 in bushels, bales, or crates, as the case may be. A chemist in anal- 

 yzing a mineral or soil or water or other substance does not content 

 himself with making a list of the elements which it contains, but 

 weighs each component carefully and expresses the results in per- 

 centages or in some other appropriate manner.* 



In this report the vegetation is discussed quantitatively through- 

 out. In each regional description the names of all the plants ol> 

 served (except the rarer ones, which are omitted) are brought to- 

 gether in a single list, without separating the different habitats. A 

 detailed study of habitats or vegetation types would require much 

 more repetition of plant names and take up more space in other 

 ways, and would be far too great an undertaking for a preliminary 

 report like this. At some time in the near future it may be desir- 

 able to take up one region at a time and classify its plants by habitat ; 

 but the present study is essentially geographical, and the geographer 



*Some illustrations of the advantages of quantitative studies in correlating 

 vegetation with soil may appropriately be inserted here. 



Two neighboring townships might both contain exactly the same species 

 of pine and oak, say two or three of the former and a dozen of the latter, 

 so that qualitative lists would show no difference between them. But if 

 a forester should find that the virgin forests of one township averaged ten 

 pines and a hundred oaks to the acre, while in the other the proportions 

 were reversed, he would be justified in concluding that there was a con- 

 siderable difference in soil, for it is well known that most oaks grow in richer 

 soil than do most pines. Some valuable observations of this cliaracter were 

 made by Coville and Branner in Arkansas a quarter of a century ago, long 

 before most botanists had ever thought of studying vegetation quantitatively, 

 f See Rep. Geol. Surv. Ark. for 1888, vol. 4, pp. 246-247.) 



It is a matter of common observation in the southeastern United States, 

 if not in other parts of the world, that deciduous trees are most abundant, not 

 onl)' in species but in individuals, on clayey and silty soils, and evergreens 

 on sandy and peaty soils. But the differences between these soils are not 

 merely physical. From all the available analyses of soils in which deciduous 

 trees predominate it appears that they are all high in potassium (having 

 usually at least 0.2 per cent soluble in hydrochloric acid), liowever much they 

 may differ in calcium, phosphorus, nitrogen, and other essential constituents. 

 .A.nd all soils with less than .05 per cent of acid-soluble potassium (or .06 

 per cent of potash, as it is usually expressed in analyses) which have come 

 to the writer's knowledge are characterized by vegetation that is mostly ever- 

 green. (The converse of these statements is not always true, however, for 

 some soils well supplied with potassium support many evergreens. But in such 



