CHAP. VII. BOTANICAL GEOGRAPHY. 297 



When the sap is frozen, the cells and vessels in which 

 it is contained are ruptured, and the parts subjected 

 to such an accident die. But trees possess a resource 

 against the effects of great cold, in their roots pene- 

 trating to a depth beyond that which the frost has 

 reached. Hence they obtain a supply of caloric, which 

 is not readily carried off, because their woody layers and 

 bark are bad conductors of heat. It has been observed 

 that the internal parts of large trees retain a temperature 

 which is about equal to that of the subsoil at one half 

 the depth of their roots. 



The temperature of a tree, being always influenced by 

 that of the subsoil, will be greater than the surrounding 

 atmosphere during winter in high latitudes, and less 

 during summer in low latitudes. This is even more 

 remarkably the case than would at first be imagined, if 

 we were to refer the cooling and heating of the earth to 

 the effect of radiation alone. But it has been observed 

 by Von Buch, that the temperature of the subsoil is 

 mainly regulated by that of the surface waters, which 

 by infiltrating into the earth produce an effect far 

 greater than any which may be ascribed to the mere 

 conducting power of rocks and soils. Now, in the 

 frigid zone, no infiltration takes place during the winter, 

 when every drop of water is converted into ice or snow; 

 and consequently the mean temperature of the subsoil 

 in very high latitudes, will be somewhat higher than the 

 mean temperature of the atmosphere ; but this is not so in 

 lower latitudes, where the infiltration continues during 

 a great portion of the Avinter. On the other hand, as 

 we approach the torrid zone, where rain falls only 

 during the coolest season of the year, the mean tem- 

 perature of the subsoil will be more cooled in propor- 

 tion than in those places where it also falls during the 

 hottest weather. Hence it happens that the mean tem- 

 perature of springs throughout the central and northern 

 parts of Europe, as far as Edinburgh, are much the 

 same as the mean temperature of the air ; whilst from 

 the south of Europe to the tropic of Cancer, the difference 

 is gradually increasing in favour of the atmos phere 



