UPPER PALEOZOIC FLORAS 341 
striking floral changes, since the effects of so important a climatic 
event must have been widespread. 
If, therefore, the glacial interruption did not occur at the close 
of the Mississippian, as Bodenbender seems to see it, we may, I 
think, continue provisionally to regard it as occurring at the close 
of the Stephanian, or possibly (as would better suit many of the 
facts) as late as the middle Permian. The totally unaffected aspect 
of the topmost Stephanian flora near Mukden in Manchuria and 
in Italy, as well as of the somewhat earlier Stephanian flora in 
South Africa, would, if evidence of other kinds to the contrary were 
absent, tend somewhat strongly to prejudice in favor of the later 
date. The moderate temperature of the fairly equable “ Permo-Car- 
boniferous” climate in general may explain why the shock of the 
climatic episode was not more severely and for a longer time felt by 
the cosmopolitan Permian flora of western Europe and America. 
