PAPERS ON BOTANY 



253 





Boxxomli'n'i 



plS'l-Son xerr? Features (^v 3 [vches 



These temperature differences at first thought do not 

 seem verv great but the difference in the number of heat 

 units received by the upland and bottomland plants dur- 

 ing any time unit must be multiplied by the number of 

 time imits in the whole growing season. When we re- 

 member that a decrease in the average annual tempera- 

 ture of North America of only four degrees would prob- 

 ably be enough to bring about another glacial period we 



