14 
1X. Bag to be applied during the whole time that the flowers 
are out. The flowers of the enclosed branch to be pollinated by hand 
with pollen from a different bush. 
Note.—In experiment VIII the flowers will probably be best pollinated by taking 
and applying the pollen ona camel-hair brush, which should be carefully sterilized be- 
forehand by immersion in spirit. 
In experiment IX the flowers may be pollinated with a brush (as 
in VIII) or by shaking pollen from a branch takeu from another bush. 
In the latter case, care must be taken that no insects are shaken on to 
the branch which is to be bagged. 
In expériments V1II and IX only muslin bags will be used ag the 
idea is to exclude pollen carried by wind. For this reason care should 
be taken as far as possible to pollinate by hand when the air is calm; 
the early morning would probably be the most suitable time if pollen 
is available then. 
Care will have to be taken as far as possiblethat all branches bagged 
are free from insects at the time the bags are applied. Care must also 
be taken that the bags themselves are empty of insects when applied to 
the branches. 
“‘T have no information regarding any special time or stage at 
which the pollen grains are liberated. Possibly some of the planting 
community may be able to tell us about this ; otherwise it will require 
observation to find out the proper time to pollinate by hand in experi- 
ments VIII and IX. Some plants only liberate their pollen-grains 
within very circumscribed limits of time ; e.g., between 4 and 7 a.M. in 
the case of Hibiscus. The times may also vary according to local 
conditions.” 
(v) Controis, 
For a proper estimation of the results of the experiments, it is 
essential to have controls, i.e, branches exposed to exactly similar 
conditions except for the bagging. For each experiment, therefore, 
two similar bushes should be chosen beforehand, each having branches 
similar to one another as regards size, situation (exposure to sun and 
wind) and promise of flower ; one of each of these branches will be bag- 
ged for the experiment, the other left untouched and treated as the 
control of the first. 
(vi) Labelling 
For accurate identification, at. the time and afterwards, of the 
branches, both of those under experiment and of those used as controls, 
it is essential that they be plainly labelled. It must be remembered 
that the actual results of the experiments cannot be ascertained until 
some months afterwards when the berries are full-grown. For labels, 
I would recommend stips of smooth wood, fastened by wire on to the 
branches to which they relate. Ifthe details are boldly written on 
both sides of the labels in pencil, they should remain legible through- 
out the monsoon. Wood slips about 4 inches by 2 inches, planed 
smooth on both sides, would be of suitable size. Thirty-two of these 
will be required for each set of experiments ; 64 for each district. 
