248 
Journal o f Agricultural Research 
voi. xxx, No. a 
These data for western yellow pine 
at Savenac Nursery indicate the follow¬ 
ing conclusions: There is practically no 
hold-over germination from fall sow¬ 
ings. When sowing is done too early in 
the fall, considerable “premature” 
germination, which is lost over winter, 
may result. It would be better not to 
sow until after the first week in Sep¬ 
tember. When sowing is done too late, 
a marked drop in germinative capacity 
usually results. It would be better not 
to sow after October 15. Between the 
two extremes the results show no con¬ 
sistent fluctuations. The five-week 
period between September 7 and Octo¬ 
ber 15 appears, from our present 
knowledge, to be a safe time to sow. 
results. The effect, however, as has 
been shown, is to lower germinative 
capacity and not necessarily to affect 
the development of plants from the 
seeds which do germinate. The date 
of sowing in the spring seems early 
enough for the average season. The 
plants used, therefore, may be consid¬ 
ered representative of the two classes 
of stock to be compared. 
About April 20, 1918, this planting 
was duplicated. In each case the nec¬ 
essary precautions were taken to elimi¬ 
nate as far as possible all variables 
except the one being tested. All 
planting was done by the same man. 
The spring and fall planted rows alter¬ 
nated, and plants from spring and fall 
I'eo 
a 70 
^60 
150 
^40 
I 
■I 
TIME OF SOWING IN THE FALL 
y 
+ 
/ 
y 
’ ^ 
-5 
''s 
-- 1 
1 
( 
\ 
\ 
\ 
'''Z 
1 
GRAPH 
NO. 
SEED SOURCE 
YEAR 
SOWN 
BASIS: 
NO. SEEDS 
SOWN ON 
EACH DATE 
'n 
s 
V 
/ 
NATIONAL 
TOREST 
YEAR 
1 
2 
3 
4 
5 
BITTERROOT 
KANIKSU 
BITTERROOT 
1915 
1916 
1916 
1917 
1916 
1919 
1000 
IOOO 
2000 
2000 
2000 
1 
20 
3! 
10 
20 
10 
20 
AUG. 
SEPT. 
DATES OP SOWING 
OCT. 
NOV. 
Fig. L—Germination of western yellow pine seed at Savenac Nursery as influenced by time of sowing in 
the fall 
It has been observed that 1-2 yellow 
pine stock from fall sowings is larger, 
more unwieldy, and more expensive to 
plant than 1-2 stock from spring sow¬ 
ings. For the production of this class 
of stock, sowing should be done in the 
spring. 
FIELD-PLANTING EXPERIMENTS 
WITH STOCK FROM SPRING AND 
FALL SOWINGS 
About October 20, 1917, approxi¬ 
mately 600 plants each of 2-0 fall-sown 
and 2-0 spring-sown stock were planted 
on a westerly aspect near Savenac Nur¬ 
sery. The fall sowing had been done 
on October 23 and the spring sowing on 
May 6 from the same lot of seed. At 
that time (1915-16) it was not realized 
that this date in the fall is for the aver¬ 
age year about a week late, for best 
sown seed alternated within the rows. 
The plantation was examined five 
times in 1918 and all plants were 
grouped in an intensive 12-term classi¬ 
fication. The more significant points 
resulting from this mass of data are 
reported here. 
Six hundred plants from each lot 
would seem to be an adequate number 
to yield reliable averages for the test. 
A smaller number would often be suffi¬ 
cient for quite uniform stock. In 
order to be sure that all irregularities in 
the stock were actually compensating 
within this number, the plantation was 
divided into four blocks. Each block 
was examined and recorded separately. 
The comparisons within each block 
corresponded closely for all blocks. 
Therefore in order to simplify this 
report the tabulations are confined to 
data from the plantation as a whole. 
