246 
H. B. SIFTON 
During the first period, if records had been preserved, there is no doubt that 
we should have found the preHminary and final counts approximating more 
closely. Although Crocker and Groves, in their work mentioned above, 
took no account of the depreciation of the seeds after 25 percent of them had 
died, their results are in agreement with those here obtained. They ob- 
serve that age lowers the percentage germination and increases the time re- 
quired for sprouting. 
Deviations from the average curve shown in figure i are found in some 
individual cases. An attempt was made to connect these with the meteoro- 
logical conditions of the year and place of growth. No satisfactory results 
were obtained, as meteorological statistics for the exact localities where 
samples were gathered had not been taken. 
The variations generally are of two kinds. In very strong samples the 
tendency is for the first period to be longer, the second shorter, and the 
division between them much sharper than in average cases. For example, 
of the Nova Scotia crop of 1902, of which eight samples were investigated 
(see table i) 96 percent germinated when twelve years old. After another 
year only 70 percent were alive, and in three more years the germination was 
reduced to 9 percent — very slightly above the average for that age. In 
weaker samples the tendency is toward a flattening of the curve. The 
Prince Edward Island crop of 1900 and the Ontario crops of 1900 and 1901 
are examples of this. Extreme examples are to be seen in the Northwest 
Territories crop of 1901 and in the Quebec crop of 1900. 
No marked difference in longevity has been observed between different 
varieties. 
Oats 
One hundred and seventy-nine samples of oats were used — fifty-two 
from the 1900 crop, sixty-four from that of 1901, and 63 from that of 1902. 
They include thirty varieties. 
Th ir longevity is much greater than that of wheat, possibly owing to 
the protection of the hulls. 41 percent of the nineteen-year-old kernels 
are still alive. The longevity curve for the oats differs from that of wheat 
in two respects. The first period is longer, and the drop in the second period 
is not nearly so steep, i.e., the kernels live longer and there are more varia- 
tions in their span of life. In the year 1900, conditions seem to have been 
less favorable for oats than the normal conditions. The vitality of samples 
gathered in this year falls off much more rapidly than is the case for 1 901 
and 1902, and this circumstance makes the second division of the curve in 
figure 2 considerably steeper than it otherwise would have been. In some 
cases the 1900 oats start out with a high germination, but in a few years 
they fall below the crops of other years. Compare for example the Manitoba 
crops of 1900 and 1902, as recorded in table 2. During the growing season 
of 1900 the temperature in general was higher than in 1902 when the 
strongest oats were produced. Slower growth due to low temperature may 
