456 BOTANICAL GAZETTE — [DECEMBER 
coats intact. Such resistance may be the usual thing in seeds 
collected and kept at an even temperature during the first winter. 
But in nature the extremes of winter climate must often destroy the 
integrity of the coats of these seeds, thus admitting the oxygen 
necessary for growth. And even if the seed coat is intact, the high 
temperature often experienced during the spring and early summer 
is sufficient to germinate the upper seeds if they are near the surface 
of the ground, since temperatures between 30 and 35°C. cause 
germination of the uppers with coats intact. Thus it is seen that 
the results of ARTHUR and MAsTERMAN may both be correct under 
proper circumstances, and the discrepancy i in | their results is readily 
accounted for. ; 
Later CROCKER (4), in testing ARTHUR’s enzyme theory, dis- 
covered that the cause of delay in the upper seed lay in the seed 
coat, which, though not excluding water, restricted the supply 
of oxygen to the embryo to such an extent that growth was tempo- 
rarily suppressed. He tested many other seeds which showed 
delayed germination and found that, contrary to the usual opinion, 
in the majority of cases the coat was responsible for the delay 
by exclusion or restriction of oxygen or water. He pointed out that 
the restriction of oxygen for the embryo by the testa gave these 
seeds an exceptionally high minimum temperature for germination, 
and that, since the seed coats of uppers and lowers differed in the 
degree of restriction, two minimum temperatures exist for each 
seed, one with the coat intact, the other with the coat removed. 
He suggested that high temperatures caused germination by hasten- 
ing the diffusion of oxygen through the seed coats; but since it is 
shown in this paper that a rise of temperature decreases the amount 
of O, demanded for growth, the question as to the influence of 
temperature on diffusion in this particular case is an open one. 
High oxygen pressures brought about germination in a short 
time, but not in the usual manner. The cotyledons elongate 
sooner than the radicle, due to the thinness of the testa at the distal 
end of the seed permitting the diffusion of oxygen more readily 
over the cotyledons. 
CROCKER (5) tested seeds of water plants also, and showed that 
in many instances the delay in germination was due. to coat char- 
