464 BOTANICAL GAZETTE [DECEMBER 
in their seed coats. CROCKER found that restriction of the oxygen 
supply by the testa was the main cause of delay in germination, and 
that difference in the degree of exclusion by the testa of upper and 
lower seed accounted for the difference in delay of the two seeds. 
But this embryonic difference, which is clearly demonstrated in the 
table, acts in conjunction with the coat differences in securing a 
longer delay in the upper than in the lower seeds. 
In comparison with the seeds and organs of other plants, the 
oxygen demand of Xanthium seeds is very high. As already 
noted, most seeds will germinate with not more than a few milli- 
meters of atmosphere, and some germinate without free O,, but 
Xanthium requires g-12 mm. of oxygen, the equivalent of 44-60 
mm. of atmosphere. This high demand for oxygen aids in secur- 
ing delay, for if only a small fraction ofa millimeter were needed 
for germination, the testas might not restrict the supply sufficiently 
to cause delay. And the difference in demand of the two seeds 
would secure a longer delay for the uppers than for the lowers, 
even if the testas did not differ in their power to exclude oxygen 
from the embryos as they do. : 
TEMPERATURE AND THE OXYGEN MINIMUM 
It was found that temperature is a powerful factor in determin- 
ing the oxygen minimum, slight changes producing marked effects 
upon the results. The potent influence of temperature in this 
regard is shown clearly by the two lots of seeds of X. pennsyl- 
vanicum kept at a pressure of 72 mm. as recorded in table I. The 
variation noted in these two lots was due to one lot being subjected 
to a temperature 6° higher than the other during a part of the last 
two days of the experiment. Previous to that time, the behavior 
of the two lots of seeds had been almost identical; but 45 per cent 
of the lowers germinated in the lot which reached 28°, as compared 
with 30 per cent, the lot which did not go above 22°; and 20 per cent 
of the uppers germinated as compared with complete failure to 
germinate at the lower temperature. The amount of growth in 
each lot shows a similar relationship. 
In- another instance shown in table I, two lots of seeds were 
subjected to atmospheric pressures of 90 and 99 mm. respectively, 
