1921] HARRINGTON—GERMINATION 355 
Discussion 
It is evident from the foregoing facts that the use of warm 
temperatures usually increases the rapidity of germination of the 
species investigated, but that comparatively low temperatures 
are more favorable for completeness of germination. In conduct- 
ing germination tests of each species, a temperature should be 
used which is warm enough to accelerate the progress of germina- 
tion as much as can safely be done. At the same time, it should 
not be warm enough to prevent the germination of any viable 
seeds, or to encourage more than is necessary the development of 
microorganisms. 
When the germination temperature is too warm, frequently the 
germinated seeds make but little growth, and it is impossible to 
judge the comparative vigor of different lots of seeds. Sometimes 
weak seeds of little value will germinate when a warm temperature 
is used, and will then appear to as good advantage as other strong 
vigorous seeds. If the germination tests are made with a more 
favorable temperature, both the strong and the weak seeds will 
germinate, but in this case the difference will be obvious at once. 
In this case some seedlings make rapid vigorous growth and are 
normal in appearance, while others have a watery translucent 
appearance, grow very slowly, and sometimes have begun to decay 
before emerging from the seed coat. On the other hand, too 
cool a temperature decreases the germination, increases the time 
required, increases also the difference in time required by different 
lots of seeds of the same species, and thus makes uniform procedure 
with the different lots impossible. This condition is well illustrated 
by larkspur samples no. 250500 and no. 250585 (fig. 9). 
In conducting germination tests of some of the kinds of seeds, 
considerable latitude is permissible in deciding upon the tempera- 
ture to be used. With certain other kinds, as larkspur, the tempera- 
ture requirements for completeness and rapidity of germination 
fall within narrow limits. 
The substratum should be such as to furnish abundant water 
to the germinating seeds without limiting the supply of oxygen. 
For this purpose folded blotting paper well moistened with water 
