CLASSIFICATION OF THE FOREST. 29 



ering which of itself operated effectually to keep out tree growth; 

 agriculture, by turning under the sod and putting crops in its stead; 

 grazing, by killing out the grass by overpasturing. The neglected field 

 and the ovcrpastured and trampled prairie then became an inviting 

 territory for the invasion of weeds and of woody growth, which 

 quickly establishes a thicket and gradually a heavy young timber 

 covering. 



It is not to be inferred that the whole Edwards Plateau is a contin- 

 uous timberland. On the contrary there is a considerable amount 

 of open grassland. Even more of it is covered with a scattered scrub 

 timber, but there is enough real timber to warrant the classification 

 of it as a timbered country. The aridity of the climate is such as to 

 exclude any such thing as a dense forest, except, as we have seen, in 

 the deep, sheltered, and watered canyons. The timber of the Edwards 

 Plateau is not one of deep shade, and of course is not accompanied by 

 all those plants and animals which love the cool twilight atmosphere 

 of dense forests. It is a dry-climate forest. The growth is stunted, 

 the Avood dense and hard, the branches rigid, the foliage somber, the 

 leaves small and stiff ; the climate is written in every feature. 



GEOGRAPHIC RELATIONS OF THE SPECIES. 



The Edwards Plateau is a common meeting ground for species from 

 the Atlantic 'forest belt, from the southern Rocky Mountains, and 

 from the north Mexican highlands. Yet the timber vegetation con- 

 tains a very considerable amount of species either limited to Texas or 

 occurring chiefly there. The situation comes about in this way: Num- 

 bers of Atlantic species, in pushing along the Coast Plain southwest- 

 ward, have encountered the more arid climate of the Southwest. 

 Surviving, they have become permanently changed into new species in 

 adapting themselves, though showing close kinship with the species 

 from their ancestral region. This circumstance was noted by Engel- 

 mann in his early articles on Texas. He says: 



* * * But shrubby species peculiar to that region represent the larger trees 

 of the same or analogous genera of the more northern parts of the country. The 

 stately walnuts of your [northeastern] forests are reduced to the low Juglans nana 

 (J. rupestris), a shrub that bears nuts the size of a musket ball. In place of the mul- 

 berry of your river bottoms, we find there the small Morus parvifolia (M. celtidifolia) 

 with leaves one-fourth the size. The fine hackberry (Celtis occidentalis) , one of the 

 largest trees on the fertile lands of our woods, is there represented by an allied genus, 

 Acanthoceltis (Celtis pallida). 



Of course wherever these species have passed on southwestward 

 beyond Texas, as numbers have done, they can not be said to be pecul- 

 iar^ Texas species. Yet it must have been in the Texas region that 

 the change from the ancestral species occurred. 



