322 
Journal of Agricultural Research voi. xxm, No. 5 
GERMINATION WITH THE DIFFERENT ALTERNATIONS 
i. Seeds which germinated equally well with all alternations 
Seeds of carrot, cucumber (Cucumis sativus L-)> muskmelon (Cucumis 
melo L.), millet (Chaetochloa italica (X.) Scribn.), onion {Allium cepa L.), 
parsley, redtop, sorghum {Holcus sorghum L.)> timothy, and watermelon 
{Citrullus vulgaris Schrad.) germinated with equal completeness with all 
of the temperature alternations from No. 2 to No. 11. They were not 
tested with alternation No. 1. 
Seeds of bean (Phaseolus vulgaris L.), beet {Beta vulgaris L-), upland 
cotton {Gossypium hirsutum L.), cowpea {Vigna sinensis (Tomer) Savi), 
flat pea {Lathyrus sativus L.), flax {Linum usitatissimun X), hemp 
{Cannabis sativus L.), meadow fescue, milo {Holcus sorghum L.), orchard 
grass, pea {Pisum sativum L.)> rice {Oryza sativa L.), perennial rye grass, 
soy bean {Soja max (L.) Piper), black bitter vetch {Vida ervilia (L.) 
Willd.) were tested only with alternations Nos. 3, 5, 6, 7, and 9 but 
germinated with equal completeness with all of these alternations. 
Some of the kinds of seed mentioned in the two preceding paragraphs 
have been shown in the earlier pages of this paper to germinate equally 
well and others almost as well at constant temperatures as with any 
alternation (see fig. 1, and p. 297). 
2. Seeds which did not germinate equally well with all alternations 
This class includes only three of the kinds of seed used in this part of 
the investigation; celery and bluegrass, which have already been shown 
strictly to require alternations of temperature for their best germination 
(see fig. 3 and 4 and text p. 299), and tomato {Lycopersicum esculentum 
Mill.). Table III shows the average percentages of germination of several 
lots of these three kinds of seed, together with a summary of the important 
features of each alternation, as shown by the time temperature curves of 
Figure 17. The germination period was 10 days for tomato, 26 days for 
Kentucky bluegrass, and 27 days for celery. 
The percentage of germination of each kind of seed showed no definite 
correlation with the extreme or mean temperatures of the alternations 
or with the shape (whether narrow and abrupt or broad and rounded) 
of the upper part of the curves in Figure 17. There was, however, a 
definite relation between percentages of germination and rate of cooling, 
as shown by the temperature at the end of 15 hours, the number of hours 
of the day during which the temperature recorded was below 22.5 0 C., 
and the number of hours at 20° — changes shown by the abrupt or gradual 
fall of the right arm of the corresponding curves in Figure 17. 
It must be admitted that the curves in Figure 17, which represent 
temperatures within blotters near the middle of the chambers, do not 
accurately represent the temperature changes occurring on top of blotters 
in the positions in which celery and bluegrass seeds were tested. It has 
already been shown, however (see fig. 13 and 17 and text, p. 320), that 
temperature differences in any one chamber in a given alternation were 
less than the characteristic differences between the different alternations. 
In all these tests the chambers were equally filled and each sample occu¬ 
pied the same position within the chamber in each test. Irregularities 
due to varying positions are therefore eliminated, and the conclusion is 
justified that an elevated temperature should not be long maintained. 
This agrees with the results of the reverse alternations discussed on 
page 305. 
