ON TURNERA ULMIFOLIA. 
115 
enclosed in large bags, which were not removed until some 
ten to twelve flowers had opened inside each. The flowers 
opened freely inside the bags and withered only a little later 
than those outside, as could easily be seen through the 
semi-transparent parchment paper. As a rule, one flower on 
each branch opened every morning, as is normally the case. 
Not a single capsule was formed, however, from any of these 
flowers, which numbered altogether over 200. 
Next five flower buds of each form were opened, and the 
stamens removed and pollen applied in each case from a 
separate plant of the same form. No capsules were 
produced. As the manipulation appeared somewhat severe 
the stamens of the pollinated flowers were not removed in 
the next experiment. Sixteen flowers of the long-styled form 
were pollinated from separate plants of the same form, and 
fourteen short-styled flowers were similarly pollinated from 
separate short-styled plants. Here again not a single capsule 
was produced : altogether forty unsuccessful pollinations. As a 
control eight long-styled and seven short-styled flowers were 
pollinated legitimately, being otherwise treated in exactly 
the same way as the last set of flowers. Seven of the long- 
styled and three of the short-styledproduced good capsules. It 
thus appears that the experimental conditions, especially the 
covering of the flowers, do not account for the entire failure 
of the illegitimately pollinated flowers to produce fruit. 
The numbers, although small, also further illustrate the 
greater fertility of the long-styled form when pollinated 
legitimately. 
Insect Visitors .—The individual flowers last only a few 
hours. The flower buds complete their.development rapidly 
during the night, but until 8 or 8.30 A.M. the corolla 
remains closely rolled up. Then, as soon as the sun begins 
to fall upon them, the corollas quickly expand. By 1 P.M. 
they are all completely withered if the sun has been bright 
during the interval. In dull weather the flowers never 
expand to quite their fullest extent. 
