Feb. 3. i 9 2 3 Use of Alternating Temperatures in Seed Germination 305 
range. In view of the temperature effects here manifested and the 
work with Johnson grass seed discussed later in this paper, it is unfortu¬ 
nate that experimental germination tests of Bermuda grass seed were 
not also made with still warmer temperatures, both constant and in 
alternation. 
ORCHARD GRASS AND PARSNIP 
These two kinds of seed required an alternation of temperatures for best 
germination. They were considerably more sensitive to temperature 
conditions, especially to the difference between constant temperatures 
and temperature alternations, than red top seed (fig. 2.), but somewhat 
less sensitive than Kentucky bluegrass (fig. 3) or celery seed (fig. 4). 
With each of these two species every alternation of temperatures except 
25 0 C. to 30° and 30° to 35 0 gave better results than any constant 
temperature. 
Reverse Alternations 
Parsnip and celery seed were tested with the alternations 25 0 C. for 
16^ hours and 20° for 7J hours each day in comparison with the alterna¬ 
tion 20 0 for i6J hours and 25 0 7J hours each day and with the similar 
alternations 30° to 20° and 20° to 30°. The alternations in which the 
cool temperature was maintained for the longer time gave uniformly 
better germination than the reverse of these alternations. The differences 
were frequently large, but with some lots of seed they were insignificant. 
JOHNSON GRASS 
It has already been pointed out (11 ): (1) That newly matured John¬ 
son grass seed is characterized by a dormancy which is not resident in 
the embryo but is imposed by coat structures and which is not entirely 
removed until after several months of dry storage, and (2) that both 
during and after the process of after-ripening the seeds of this species 
germinate better at a high temperature and with temperature alterna¬ 
tions than at lower or constant temperatures. In the earlier paper (rz) 
an alternation between 25 0 and 40° C. was given as the optimum tem¬ 
perature condition for the germination of this kind of seed, but later work 
has shown that the still higher temperature alternation 30° to 45 0 gives 
still better results. 
Figure 6 shows the average rates and percentages of germination of 
six lots of fully after-ripened Johnson grass seed under 12 temperature 
conditions, the higher temperature in each alternation being maintained 
from six to eight hours each day. 
The accelerating effect of warm temperatures is shown by the fact that, 
in contrast to all the kinds of seed hitherto discussed, Johnson grass seed 
germinated more rapidly in the early days of the germination test at 
constant temperatures than with the temperature alternations correspond¬ 
ing to a lower mean temperature, but just about the same as with tem¬ 
perature alternations corresponding to the same mean temperature. The 
high points in the graph for germination in two days all correspond to 
constant temperatures, but the points in this graph showing germination 
with the temperature alternations 20° to 35 0 C., 25^040°,and30^045°, 
fall on almost exactly the same levels, respectively, as those for the 
approximately equivalent constant temperatures 25 0 , 30°, and 35 0 . 
