478 BOTANICAL GAZETTE [JUNE 
The most interesting difference between the two types is shown 
also in these figures. The photographs were taken at the same 
time, about September 1, 1913. Type II had shed all of its pollen 
and its burs were full grown. One shriveled cluster of staminate 
flowers can be seen above; while type I is just beginning to open its 
first staminate flowers, and its carpellate flowers are almost too 
small to be seen. Type III was intermediate, having about half 
| 
| 
| 
a4 nae asdnnad. 
FO UAL 
WPAN OUND Me 
2 ee ee 
Fah OO anne 
Fic. 4.—a and b, uppers and lowers of X. globosum Shull; c and d, uppers and 
lowers of X. pennsylvanicum Wallr.; e and f, uppers and lowers of X. canadense Mill.; 
all natural size. 
grown burs at the time type I sheds its pollen; its flowers had no 
doubt been generally pollinated from plants of the same type. In 
this interesting condition we find the explanation of the remarkable 
uniformity exhibited by the offspring of plants taken at random 
from a heterogeneous population in the field. There is a physio- 
logical isolation that effectually prevents hybridization in the great 
majority of cases. The pollen of one variety has been shed long be- 
fore the stigmas of the other are ready for the pollination processes. 
