898 
Journal of Agricultural Research 
Vol. XXXI, No. 9 
Table V .—Maturity test: Germination according to time of ripening 
Collection date 
Greenhouse germination 
Weight 
of 1,000 
clean 
seeds 
50 days 
100 days 
150 days 
210 days 
Aug. 6___ 
Per cent 
18.8 
7.0 
1.7 
1.3 
Per cent 
58.7 
48.7 
11.7 
5.0 
Per cent 
61.2 
69.2 
58.8 
49.0 
Per cent 
61.2 
70.7 
78.3 
80.0 
Grams 
11.40 
13.35 
18.50 
17.35 
Aug. 15___ 
Sept. 1... 
Sept. 17 *__ _ . 
* Seed falling after this date. 
This more rapid germination rate of the earlier seed may have 
been caused by immature seed coat and the greater total germina¬ 
tion of the later collection by a greater number of seed attaining 
ripeness. 
When it is now considered that the sample in this maturity test 
collected September 1 gave a final germination of 78.3 per cent, 
which was as high as that in the after-ripening test collected Septem¬ 
ber 4, there seems little ground for supposing that the seed was 
unripe or green except perhaps as regards the seed coat. 
Sowing of fresh seed in the autumn will not always bring prompt 
spring germination, for if much rainy winter weatner occurs, with 
light snow and temperatures of from 32° and 50° F., the seed will be 
injured. 
Observations on the character of cones and seed showed that on 
on August 6 the cones were purple and the seed light brown, not 
filled out, and somewhat muky. On August 15 the cones were 
still purple and the seed light-colored but firmer and better filled. 
On September 1 there were also many light-colored seed, but all 
were filled out as well as those of September 17. The appearance of 
the seed and the weight on September 1 indicate that it was ripe at 
this date. 
A microscopic examination was made of the structure and com¬ 
position of the tegumentary system of the seed collected August 6 
and September 17 from the same tree for the purpose of finding out 
whether the differences in time and amount of germination were due 
to greater development or hardening of the seed coat after August 6. 
In Doth samples the identical structure and composition were ob¬ 
served both by measurements and microchemical tests of the seed coat 
and peridermium thicknesses. It was apparent in cutting the sec¬ 
tions, however, that the seed last collected had a harder coat. By 
soaking in concentrated sulphuric acid, the carpels of the seed collected 
August 6 decomposed sooner. The tests showed that the composi¬ 
tion of the main hard seed-coat layer in both cases had not pro¬ 
gressed beyond the formation of cellulose and that the cells of endo¬ 
sperm and embryo of both samples were filled with proteid. It 
was not necessary to use the microscope to determine that the 
endosperm of the early seed did not fill the carpel as well as that of the 
seed collected later. 
Before concluding, it should be emphasized that a higher quality 
of seed has been produced from year to year by avoiding too high 
temperatures during extraction, and molding in storage. Investiga¬ 
tions with this and the other species, as well as experiments con- 
