448 
Joumal of Agricultural Research 
Vol. XXVIII, No. 5* 
The lights were switched on before sunset and off at the number of hours, 
after sunrise requisite for making the desired light exposure or day length. All 
of the plants in the A series were given a 10-hour daily exposure at first. The 
day length was later modified in different ways until the autumnal equinox,. 
1923, from which time all the plants of the series were exposed to normal day 
length until the close of the period of observation, November 30, 1923. The 
following were the modifications made in the 10-hour daily period: 
Plants in cans No. 1 to 18, inclusive, A series, were divided into groups 1, 2 r 
and 3, November 10, 1922. 
Group 1: Plants in cans No. 1, 4, 7, 10, 13, and 16 were subjected to a 10-hour 
day as before until March 14, 1923, when the day was lengthened to 12 hours. 
Group 2: Plants in cans No. 2, 5, 8, 11, 14, and 17, were subjected to a 12?- 
hour day for 10 weeks, that is until January 19, 1923, and then to a 12-hour day. 
Group 3: Plants in cans No. 3, 6, 9, 12, 15, and 18 were subjected to a 13i-hour 
day for 10 weeks, that is until January 19, 1923, and thereafter for 4 weeks to a, 
13-hour day, then for 2 weeks to a 12i-hour day, and finally, to a 12-hour day- 
The plants in the remaining cans of the A series, No. 19 to 27, inclusive, were 
subjected to an 18-hour day length the first week after March 14, 1923, and the 
length was thereafter reduced by one-half hour weekly until May 17, after which 
it remained at 13i hours for 2 weeks, and then was further reduced to 13.2 hours, 
the normal day length, for 1 week. After June 6, the cans of this lot were divided 
into groups 4 and 5. 
Group 4: Plants in cans No. 20, 22, 24, and 26 were subjected to normal 
day length. 
Group 5: Plants in Cans No. 19, 21, 23, 25, and 27 were subjected to a 12-hour 
day length. 
When subjected to the normal day length, the plants in cans flowered during 
both seasons a little later than did those in the field. Flowering and fruiting 
were normal, otherwise. 
In cans No. 1 to 13; inclusive, B series, one or more plants came into blossom 
within a period of two weeks, October 11 to 25, 1922, and all but four which 
opened their first flowers in November had blossomed by the end of October. 
(PI. 3, A.) Plants in cans No. 14 to 27, inclusive, B series, like those in the 
field tests, produced no flowers, all planted as late as June 21 failing to bloom. 
The plants in cans No. 1 to 13, inclusive, A series, which had been given a 
10-hour day throughout this period, showed no indication whatever of budding. 
(PI. 3, B.) 
By October, 1923, of the 39 plants in cans No. 1 to 13, inclusive, B series, 
2 had died and 3 had made a very poor weak growth. Of the remaining 34 
plants, 22 came into their second blossoming in October, and 6 in November. 
The regularity in effect of the recurrent annual factor which induces blossom¬ 
ing may be noted by a comparison of the dates on which first blossoms appeared 
in cans No. 1 to 13, inclusive, B series, during two seasons, as shown in Table I. 
Plants in cans No. 14 to 27, inclusive, B series, produced their first blossoms 
in October, 1923, when 31 of the 41 still living plants came into blossom. In 
November, 6 of the remaining 10 blossomed, 2 showed well-defined buds, 1 died 
budded, and 1 which had tasseled in October failed to develop. 
That the interval between the first indication of budding, the appearance of 
the terminal tassel, and the opening of the flowers is closely related to the day 
length, was shown by observations made on the dates of the first budding and of 
the opening of the flowers. The plants in cans were carefully examined at weekly 
intervals for the first appearance of tassels. Table II gives the dates on which 
the tassels were first observed on 20 plants in cans No. 19 to 27, inclusive, B series, 
and the dates on which each opened its first flower. 
