340 PLANT SOCIOLOGY 



and derived from various communities which have developed at no 

 great distance from the bog. If we find in a hmited area several pollen 

 spectra with synchronous horizons, we can determine with some 

 certainty the sequence of forest conditions, at least in so far as the tree 

 layer is concerned. The brilliant investigations of Von Post and 

 Erdtman in Sweden, of P. Stark, Bertsch, Overbech, Firbas and 

 others in Germany, of P. Keller, Furrer, Liidi, and others in Switzer- 

 land suffice to establish for northwestern Europe the post-glacial 

 forest periods (Woodhead, 1928) as shown on page 339. 



Pollen analysis gives good bases for conclusions about the chrono- 

 logical sequence of the terminal societies, the forest climaxes, and 

 therefore about changes in the general climate of extended areas. Of 

 course, a single pollen diagram is not sufficient. When by a more or 

 less dense network of investigations a chronological correspondence has 

 been established between the forest periods determined by pollen 

 analysis, then only may we assume changes of climate as the dynamic 

 cause of the changes of forest.^ The explanation of changes of chmate 

 should involve an evaluation of all the findings of related sciences (see 

 also Liidi, 1930). The above attempt to depict the forest periods of 

 eastern Europe represents a synthesis of available prehistoric data 

 (Table 39). 



1 Even then, as pointed out by Furrer (1927), the differing mobility of tree 

 species by seed distribution must be taken into account. 



