THE PAIRED SEEDS OF COCKLEBUR. 23 I 



detect and compare corresponding seeds in each, and further study 

 of the subject was taken up with renewed interest. 



With this new information at hand a lot of cockleburs were 

 placed in wet sand to germinate. Out of 223 from which one 

 seedling appeared, it was the lower seed which grew, except in 

 fifteen burs. Upon cutting open the fifteen exceptions, eleven 

 were found to contain no viable lower seed, and should conse- 

 quently be excluded. This trial, therefore, gave ninety-eight per 

 cent, germination for the lower seed against two per cent, for the 

 upper. The result was in general confirmed by further observa- 

 tions of the same nature. 



Comparison of the weights of the upper and lower seeds was 

 now undertaken ; when the lower seed exceeded fifty milligrams 

 the upper proved to be lighter, but when the lower seed fell below 

 fifty milligrams the upper was heavier. It is clear that when the 

 burs are well developed, the lower seed will be larger and stronger 

 than the other, and is the one that will grow quicker. This was 

 found to be true even when the seeds were removed from the bur 

 and placed in a germinating pan. A number of weighed seeds 

 were thus tested, and after three to four months twenty per cent. 

 of the lower and also heavier seeds remained in viable condition, 

 and over ninety-two per cent, of the upper and lighter seeds. A 

 number of trials of this sort were made with fairly uniform 

 results. 



Comparing the various observations, both those where accurate 

 data were recorded and others less complete and exact, the general 

 statement appears to be warranted, that as a rule the lower seed 

 in a cocklebur is better developed, sensibly heavier, and will grow 

 sooner than the upper one. If burs are planted in the soil, the 

 lower seed usually grows the first year after maturity, and the 

 upper seed grows the second or some subsequent year. When 

 removed from the bur and placed under conditions for germina- 

 tion, both upper and lower seeds grow more readily than when in 

 the bur, but still dififer considerably, the lower growing sooner 

 than the upper. 



This rule appears to hold good for all species of Xanthium. 

 The one most employed in the above experiments was X. Pcnnsyl- 



