136 SOUTH AFRICAN BOTANY 
again. Place itin the sunlight for a few hours, then pick 
a third leaf and test it. 
Resuits.—The second leaf shows no purple colour with 
iodine and therefore contains no starch, the first and 
third leaves both turn purple. 
Deduction.—The plant loses its starch in the dark, 
but makes more when again exposed to the light. 
Method (i1.).—Pin pieces of silver paper across a leaf 
on a plant, so as to block out the light from part of the 
leaf; they must not be pinned tightly enough to exclude 
air. After two days pick the leaves and test for iodine. 
Tt will be found that only the parts of the leaf which 
had light contain starch, the others contain none. 
105. Experiment 13.—To see if carbon di-oxide is 
necessary for starch formation. 
Apparatus.—Caustic potash, bell-jar, dish, leaves. 
Method.—Set some leaves with their stalks in a glass 
of water, stand them beside a dish of caustic potash 
solution, and cover the whole with a bell-jar. Set the 
apparatus in a good light, and leave for two days. Fix 
up another experiment as a control in which the con- 
ditions are exactly the same, except for the dish of 
caustic soda (fig. 75). This substance absorbs carbon 
di-oxide, hence the leaves under the first jar have none 
of this gas, those under the second have. At the end 
of two days boil and decolorize both sets of leaves and 
test for starch. 
Result.—Those from the first jar de not contain starch, 
those from the second do. 
Deduction.—Carbon di-oxide is necessary for starch- 
making. 
