18 The: Colorado Experiment Station 
that an increase of 15 plants had taken place, consisting apparent¬ 
ly of new seedlings together with the recovery of most of those 
first dug out. On May 10, 1917, there were 57 plants on the plot, 
including the smallest, all of which were somewhat inferior in size 
and vigor to those outside this area. This denotes a total reduc¬ 
tion of about 47 per cent from the two diggings. 
Plat II consisted of a strip 5x38 feet in a lawn from which the 
dandelions had been dug during the preceding May. The plants 
on this area, 302 in all, were dug again on September 23. 
The area, when inspected May 15, 1917, showed a decrease 
of 105 (36) per cent, and those left were small and weak, with about 
Fig. 9. (a) Four florets in different stages of blooming, taken from the 
same flower head, (b) Mature “seeds” with parachute of hairs. Twice natural 
size. 
10 per cent able to produce bloom. The results of these two ex¬ 
periments seem to indicate the importance of digging in late sum¬ 
mer and autumn, in addition to that in spring. The first or early 
digging prevents the ripening of seeds by the plant, while the 
later digging weakens them by causing new growth, at the expense 
of reserve food, late in the season, or kills them outright. 
Plat III covered 190 square feet in the same lawn as Plat II. 
This area was sprayed September 23 with an 8-percent solution of 
iron sulphate. Inspection of this plat on May 15, 1917, showed an 
increased number of plants amounting to nearly 19 percent, con¬ 
sisting, evidently, of new seedlings. From this experiment it ap¬ 
pears that one weak spraying is of little effect in controlling the 
weed, even when used late in the season. 
Plat IV covered an area of about 120 square feet in a different 
location on the same lawn as the preceding plat. This was given 
