26 
BULLETIN 57, HAWAII EXPERIMENT STATION 
June 6 : New top growth observed, also new leaves at apex of even the oldest 
stalks. The new tops are mostly of young Group 3a. A long break between the 
new and old stalks, which latter belong mostly to old Group 2. Some new root- 
stocks growing. New top growth is developing from old rootstocks growing in 
February. 
August 8: Hard winds of past two months blew down many old stalks. Field 
looks like a new one. The new tops average medium to old Group 3a. New 
rootstocks apparent, but small. 
November 17: A profusion of new spikes developing with little or no new top 
growth. 
February, 1926: Very little growth of either stalks or rootstocks since Novem- 
ber. Hard winds (kona) flattened most of the existing top growth. Many 
spikes beginning to develop,. New top growth expected. 
Table 12 gives the monthly yields and tare of the crop. 
Table 12. — Monthly harvests and tare of one and two seed plantings from 0.1-acre 
plats 
Age of 
crop in 
.months 
Month of harvesting 
1925: 
April 
May 
June 
July 
August 
September 
October. .. 
November 
December. 
1926: 
January... 
February . 
Tare 
Per cent 
12.7 
12.7 
11.0 
15.5 
12.1 
11.4 
15.3 
8.9 
8.0 
9.8 
One-seed planting 
Gross 
yield 
per acre 
Tons 
15.09 
17.25 
18.75 
18.60 
23.20 
29.85 
32.90 
38.20 
40.75 
44.93 
47.60 
Net 
yield 
per acre 
Tons 
13.17 
15.06 
16.69 
2 16. 37 
20.39 
26.45 
2 28. 95 
34.80 
37.51 
40.48 
42.93 
Monthly 
incre- 
ment 
Tons 
1.89 
1.63 
3 3.70 
6.06 
2.50 
5.85 
2.71 
2.97 
2.45 
Two-seed planting 
Gross 
yield 
per acre 
Tons 
21.70 
26.50 
28.58 
27.82 
30.25 
33.23 
37.00 
40.90 
44.25 
46.55 
48.70 
Net 
yield 
per acre 
Tons 
18.94 
23.13 
25.44 
2 24.48 
26.59 
29.44 
3 32. 56 
37.26 
40.71 
41.94 
43.93 
Monthly 
incre- 
ment 
Tons 
4.19 
2.31 
3 1.15 
2.85 
3.12 
4.70 
3.45 
1.23 
1.99 
1 Tare was ascertained monthly by determining the proportion of adhering soil, roots, dead scales, and 
^excess stems on about 200 pounds of rootstocks. (See also "Tare," p. 34.) • 
J The very high tare values, together with actual decreases in yields in two instances, brought into question 
the accuracy of the tare values. Therefore 12 per cent, the average tare of the other months up to Novem* 
ber, was used to compute the tare weights. 
* Increment based on June harvest since the yields in July were less than in June. 
Table 12 shows a nearly continuous though irregular growth of 
canna from 9 to 19 months, inclusive. The rootstocks increased 
both in number and in weight. From April to July, inclusive, the 
rate of growth constantly decreased. Beginning with August, 
growth was resumed, reaching its maximum in November, when it 
again decreased. 
For the first 12 months of growth (until August, 1925) the two-seed 
planting was decidedly superior to the one-seed planting. The 
July harvest of the former yielded 24.48 tons per acre net, whereas 
that of the latter yielded only 16.37 tons. Beginning with the 
thirteenth month the one-seed planting grew more rapidly than the 
two-seed planting and at 19 months it lacked only a ton of equaling 
the latter. These results are somewhat different from those given in 
Table 10, where the two-seed plantings are shown to have exceeded 
the one-seed plantings by 6.1 and 5.6 tons, respectively. The large 
monthly increments of the one-seed planting in the last seven months 
of growth may have been due partly to differences in seed or fertility 
of soil. 
Sixteen hills were selected at regular intervals from each of the 
plats and classified according to the method outlined under " Classifi- 
cation/ ' page 12. Table 13 gives the results. 
