CARBOHYDRATE METABOLISM AND GROWTH IX EDIBLE CANNA 5 
METHODS OF INVESTIGATION 
FIELD METHODS 
Two general methods were devised for continuous study of the 
growth of edible canna. 
In the first method, the different generations of the rootstock were 
studied, tracing the progressive growth of the hill from the original 
seed rootstock. The hill was carefully dug and the individual 
plants were separated and grouped according to successive genera- 
tions, beginning with the plant developing from the original seed 
rootstock as the first generation. Notes were then made of the 
number, kind, and size of the rootstocks of each generation, the 
Fig. 4.— Top view of a hi!l ol canna during its optimum growing stage (4 months old). The seven 
large spike rootstocks, all of the fourth generation, were supported by nine vigorous stems. Note 
the profusion of smaller secondary "spikes " developing in the center of the hill 
stage of maturity of stems, and the amount of both primary and 
secondary growth. 
In the second method, the classification of rootstocks was studied, 
the lot from each hill being grouped as dormant, mature, and imma- 
ture. 
The procedure followed was to dig the hill of canna and group 
the different plants in the hill according to the stage of maturity 
of their stems. Group 1, dormant stage, composed plants on which 
the leaves had died and the stem had partly or wholly shriveled. 
This group appears when the hill is 8 to 12 months old. Group 2, 
mature stage, comprised plants whose new growth within the stem 
had ceased, and, in case of the older members, the lower leaves had 
shriveled. Group 3, immature stage, comprised plants the stems of 
which had not attained their maximum growth. Group 3 was sub- 
divided into two lots (a) plants with stems developing, comprising 
