LONGEVITY OF THE SEEDS OF CEREALS, CLOVERS, AND TIMOTHY 245 
to flatten again. Its direction in the seventeenth year is practically parallel 
to that in the tenth, and in the eighteenth and nineteenth years its direction 
corresponds with that in the fifth and sixth years. 
When an average sample of wheat is stored for a long period, this in- 
vestigation shows that its depreciation is divided into three more or less 
distinct periods. During the first few years (ten or eleven on the average 
for our Canadian samples), the weak grains gradually die. After this 
comes the period when the seeds of average vitality (forming the bulk of the 
sample) die very rapidly. A few seeds, very tenacious of life, are still 
left, and slowly lose their vitality during the 
final period of about three years. 
It would be interesting for some one who 
has proper facilities at his disposal to raise 
plants from the seeds of these three divisions 
separately, and determine whether there are 
separate genotypes with different powers of 
resistance and whether their hardiness is trans- 
mitted to the plants which they produce and 
to their descendants. The work of Crocker 
and Groves (Proceedings of the National 
Academy of Sciences, March, 1915) tends to 
show that the loss of viability is due to the 
coagulation of the cell proteins, and if this is 
so it would be unnecessary to wait for years 
for the seeds to die at ordinary temperatures. 
Their vitality could be destroyed by carefully 
regulated heat after the manner used by these 
investigators. 
The results so far considered have to do 
with total germination percentages as deter- 
mined by a ten-day test in a standard germin- 
ator at alternating temperatures of 20° C. to 
30° C. Unfortunately complete records of the 
preliminary four-day tests, which are believed 
to give an indication of the energy of germination, are not available. Records 
of these were kept after the tenth year, however, and are recorded on the 
graph by means of the dotted curve, which follows a course practically paral- 
lel to that of the main curve. During the second period of depreciation the 
preliminary count is from 6.5 to 8 percent lower than the final count. 
The energy of germination of a considerable portion of the seeds which are 
to die during the next year has been weakened. During the third period 
there is a smaller absolute difference between the two results, but a greater 
comparative difference; i.e., of the seeds left alive after the end of the second 
period, the greater proportion have lost most of their energy of germination. 
i 
Fig. 
. Longevity curve for 
spring wheat. 
