8 BULLETIN 542, U. S. DEPARTMENT OF AGRICULTURE. 
After a heavy dew they will be found in this swollen condition, 
but if not washed off the anther they will be found later in the day, 
if the sun has come out, to have resumed their normal dry form? 
oblong-oval. 
THE PISTIL. 
It has been suggested that some of the unproductive varieties 
might be found to have abortive pistils and hence to be incapable of 
being properly fertilized. For this reason the pistils of every avail- 
able variety were examined. The only difference noticeable among 
most of them was a slight variation in the width or expansion of the 
stigma ; in some varieties the stigma is not expanded at all, being no 
wider than the style below it; in a few others it is commonly slightly 
expanded, and flattened on the upper surface. A V-shaped cleft is 
frequently visible in the outer edge of the stigma, extending down- 
ward in the form of a gradually narrowing groove and finally disap- 
pearing. In the freshly opened flower the pistil stands nearly erect, 
but the stigmatic surface is often slightly to one side and facing away 
from the center of the flower. Frequently the upper end of the style 
is curved slightly in this direction, especially when the flower com- 
mences to get old. 
On one tree only were the pistils found to be commonly abortive ; 
this was a Mulgoba seedling on the Hugh Matheson place at Cocoanut 
Grove. On this tree many of the pistils were exceedingly short or 
almost entirely aborted, the rudiments of the style persisting in the 
form of a short point on the upper surface of the ovary. Some of the 
pistils seemed to be normal, but a great many were in this abortive 
state and certainly incapable of being fertilized. 
PROTECTION OF POLLEN FROM DEWS AND RAINS. 
A frame was erected over one side of a young tree of the Julie 
variety and a canvas curtain placed over it every afternoon about 
4.30 o'clock and removed in the morning soon after sunrise. This 
curtain was also put over the tree whenever a shower threatened. 
One-half of the tree was left exposed to the elements as a check. 
It was thought that by protecting the flowers from rain, so the 
pollen would not be washed away, there might be a better chance of 
pollination. By carefully comparing the flowers on the covered and 
uncovered sides of the tree on days when the weather was fair or 
showery or there had been a heavy dew the previous night, it became 
evident that the protection afforded could not be of much practical 
benefit. The covering did not prevent the access of sufficient moisture 
to cause the pollen grains to swell and become sticky, and although 
it prevented the pollen from being entirely washed away by a 
shower it did not seem to have a great deal of bearing upon the aver- 
age of pollination. The cover was kept on for several weeks and then 
