POLLINATION OF THE MANGO. 13 
difficult of all to germinate. The percentage of grains which sent 
out pollen tubes varied from to a little more than 50, but in the 
vast majority of the cultures would not average above 10 or 15 per 
cent. No pollen tubes developed to a length exceeding seven times 
the diameter of the pollen grain, and commonly they did not exceed 
four or five diameters. Without going into detail regarding the 
cultures, of which about 300 were made, it can be said that the 
germination of all varieties was very poor. While the percentage 
of germination was higher in some varieties than in others, it varied 
in the same variety upon different days, though the conditions were 
as nearly similar with regard to culture medium and temperature as 
it was possible to keep them. The percentage of germination and 
the vigor of the pollen tubes were greatly inferior to those obtained 
in cultures of the feijoa (Feijoa sellowiana) , the lumbang (Aleurites 
moluccana), or several other plants which were tested. It has been 
pointed out by Prof. E. J. Kraus, however, that pollen germinations 
made under moist-chamber conditions do not necessarily indicate 
the germinability of the grains on the stigmas under optimum 
conditions. 
THE PRODUCTION OF FRUIT. 
The stigmas, as has been shown, receive no pollen in more than 
half the instances; and of those cases where they are pollinated, in 
comparatively few are there more than one or two grains upon each 
stigma. 
Under these conditions it was thought that the production of 
fruit might be increased noticeably if the stigmas were pollinated 
more abundantly ; with ample pollination, even if some of the pollen 
grains were defective, a few capable of developing pollen tubes and 
reaching the ovum should reach each stigma. Working on this 
theory, a number of hand pollinations were made. 
On a 6-year-old tree of the Bennett variety at the E. B. Douglas 
place, Buena Vista, Fla., five panicles of flowers were selected for 
hand pollination, and on two days, April 28 and May 1, 1915, pollen 
from the same tree was applied to all the fresh stigmas on each pan- 
icle. The exact number pollinated on each panicle was as follows: 
Panicle No. 1 30 stigmas. 
Panicle No. 2 . 25 stigmas. 
Panicle No. 3 31 stigmas. 
Panicle No. 4 25 stigmas. 
Panicle No. 5 51 stigmas. 
These were not all of the perfect flowers on each panicle, but it was 
thought advisable to pollinate only those stigmas which appeared 
to be fresh and in a receptive condition at the time the work was 
done. No fruits were carried on any of these panicles, though on 
May 21 about 80 well-formed fruits were counted on other panicles 
