to The Colorado Experiment Station 
In another experiment, made to determine how late in the sea¬ 
son the plants can produce viable seed, 470 seeds were taken at 
random from several different plants on November 29. At this 
time the ground was frozen quite hard, although open, dry weather 
prevailed. Some of the blossom heads were not fully matured, 
which probably accounts for the low percentage of germination, 
16 per cent at this time. It is evident, however, that some good 
seeds are being formed as late in the season as it is possible for 
blossoming to occur, and although these seeds are too late for 
germination at once, they will be on hand for early sprouting the 
next spring. 
Fig 1 . 6. How the dandelion root divides as it grows older. Left hand root one 
year old, right hand root three or more years old 
In order to find out how soon after blossoming the flower 
heads were able to mature seed capable of germination when re¬ 
moved from the plant, the following experiment was carried on: 
Eighteen flower heads, just opening for the first time, were 
picked and allowed to lie on the ground until wilted, after which 
they were taken indoors and allowed to become dry. Examination 
later showed that in no case had any seeds matured, although some 
of them had that appearance. 
Thirty-six flower heads were marked with numbered slips of 
paper and careful records of them were kept during ten days. 
