i64 FORMATIONS AND GUILDS [Part II 



To consider grassland, as is frequently done, as the sign of a 'bad 

 climate,' as an evidence of poverty in Nature, as a transition state between 

 forest and desert, is at best comprehensible from a forester's point ol 

 view, but is neither scientifically nor practically justifiable. Indeed certair 

 forms of woodland are climatically more accommodating in their demand; 

 than is grassland. Victory in the struggle between woodland and grass-i 

 land belongs to the one of the two antagonists with which the givei 

 climatic conditions best correspond. 



Accurate knowledge of the demands made on the one hand b\- wood) 

 plants, and on the other by grasses, in regard to atmospheric precipita 

 tion, movements of the air, and heat, will yield us the elements of whicl 

 a woodland climate and a grassland climate are composed. 



ii. WOODLAND CLIMATE. 



Let us first consider the woody plant in its most complete development 

 as a tree. In the tree the transpiring surface is at a greater distanc 

 from the water-supply in the soil than it is in the shrub or herb ; besidi 

 this, the strata of air surrounding that transpiring surface have propertic 

 different to a certain extent from those nearer the soil ; finally, at leas 

 in many cases, the transpiring surface of the tree is larger when compare^ 

 with the corresponding surface of the ground than it is in the shrub o 

 herb. 



On the other hand, the tree has at its disposal a vast root-s)'sten 

 which is capable of utilizing very deep-lying supplies of water, and upo 

 these it often mainly relies, as its root-tips for the most part lie ; 

 a considerable depth below the surface of the ground. 



Our present knowledge of the physiology of trees is in the mai 

 derived from the Central European flora, the trees of which all transpiii 

 freely although in an unequal degree, and are correspondingly highll 

 water-demanding and hygrophilous during the vegetative season. Tl' 

 most comprehensive and useful investigations regarding the transpiratio 

 of the forest trees of Central Europe have been carried out by R. vc 

 Hohnel \ from whose works the following data are taken : — 



The author experimented with seedling-trees 5-6 years old and 50-80 cm. Iiil 

 which had been planted in ordinary garden-pots 16 cm. high and each containii 

 3J-5 kilograms of earth. Tiie pots were surrounded by wide air-tiglit sheet-zii 

 cases, made so that not only was it possible to water the plants while the p 

 was kept completely enclosed, but also that the soil did not come into ai 

 contact with the zinc. In this way all loss of water from the soil was exclude 

 and a correct determination of the amount of transpiration secured. Tl 

 experiment was also so arranged that the pots could not be exposed to tl 



' Von Hohnel, op. cit. 



