324 
Journal of Agricultural Research 
Vol. XXIII, No. s 
was most rapid with alternation No. 9 and least rapid with alternation 
No. 4, although the upper temperature extreme was practically the same 
in the two alternations. 
The germination of the seeds of the cucurbits was especially accelerated 
by the warm alternations. The average percentage of germination of 
watermelon seed in three days was 66 with alternation No. 9 and only 
6 with alternation No. 4. 
4. Uniformity of results with favorable alternations 
Some lots of seed are much more, sensitive to temperature conditions 
than other lots of the same kind. Consequently, while the most favorable 
temperature conditions give with every lot approximately the maximum 
germination attainable, less favorable conditions (for general use) may 
give equally good results with some lots and very poor results with 
others. A really very good, but temperature-sensitive, lot of seed may 
thus under slightly unfavorable temperature conditions germinate more 
poorly than a very poor lot which is not markedly sensitive to temperature 
conditions. It is impossible to say just what constitutes this different 
temperature sensitiveness of different lots of the same kind of seed, but 
it frequently becomes evident. It was strikingly exemplified by the 
celery seed used in this investigation. This suggests the possibility 
that some of the kinds of seed which in this investigation germinated as 
well with constant as with alternating temperatures may, under some 
circumstances—for instance, incomplete after-ripening—germinate better 
with an alternation. 
Figure 18 shows the germination of eight lots of celery seed with 
four temperature alternations Nos. 2, 3, and 4, which were very favor¬ 
able alternations, and No. 11, which was unfavorable. The curve for 
alternation No. 5 would follow almost exactly the course of that for 
alternation No. 2. Celery seed No. 87602, though one of the best ger- 
minators under favorable conditions, germinated less than half as much 
as the poorest lot when tested with alternation No. 11. 
The difference in sensitiveness of different lots to temperature con¬ 
ditions is even more strikingly shown in Figure 19, which shows the 
germination of three lots of celery seed with 10 temperature alternations. 
COMPARISON WITH THE RESULT OP PlELD TESTS 6 
Temperature changes in the soil during the spring and early summer 
Several years ago, in connection with a series of field tests of various 
kinds of seed, records were kept of the temperature of the soil at various 
depths on Potomac Flats, near Washington, D. C. The period covered 
by the tests was from March 26 to June 28, inclusive. Nearly every 
day during this period temperature observations were taken at intervals 
of from 15 to 30 minutes, beginning at 4.30 a. m. and continuing to 
about 10.30 p. m. 
From March 26 to April 22 the temperature of the soil at a depth 
of 1 inch was always below 30? C. and on several different days did not 
not become as warm as 20°. The period from April 22 to May 1 was 
somewhat warmer, but the temperature of the soil at 1 inch was 
never above 31 0 . and on one day barely reached 20°. On 18 of the 
28 days in May for which there are records the daily range was from 
20 0 or cooler to 30° or warmer. The other 10 days for which there are 
records were cooler. From June 1 to 7 and from June 11 to 15 the 
• The data for soil temperatures and seedling production used in this comparison are from unpublished 
notes furnished by Mr. E, Brown, botanist in charge of the seed-testing laboratories. 
