tgtt] SHULL—OXYGEN MINIMUM AND GERMINATION 473 
testa has been removed, and those obtained by Crocker with the 
same kind of seeds with the testa intact. When the seeds with 
testas on are allowed to germinate normally, the hypocotyl always 
elongates first and the root is well developed before the cotyledons 
begin to grow. That is, there is a strong correlation between the 
growth of these two organs of the seed. But if the well ripened 
seeds with coats on are germinated in an atmosphere containing 
a high percentage of oxygen, the cotyledons instead of the hypo- 
cotyl initiate the development of the embryo. The strong normal 
correlation is reversed by temporary suppression of the hypocotyl, 
due probably to a more rapid diffusion of oxygen through the very 
thin distal portion of the testa surrounding the cotyledons than 
through the thick proximal portion which invests the hypocotyl. 
The cotyledons thus receive the necessary free oxygen for germina- 
tion sooner than the more sensitive hypocotyl, with the result that 
the usual course of growth in the seed is completely changed. 
This reversal of a very strong correlation probably requires 
considerably more oxygen than a normal germination. Not only 
is the usual behavior overcome, but the part which grows instead 
of the hypocotyl is a much less sensitive part of the embryo, 
undoubtedly requiring more oxygen to initiate its activity. In 
animal cells food storage renders the protoplasm inert, and delays 
cell division to a very marked degree. It is possible that food 
storage in the cotyledons has a similar effect on their activity, 
requiring more energy, and therefore more oxygen, for the growth 
processes. 
Another factor probably responsible for a considerable portion 
of the gaseous exchange during germination with the seed coats 
intact is the testa itself. BECQUEREL (2) has shown that the 
integuments of seeds produce CO, in comparatively large amounts, 
sometimes greatly exceeding the seeds from which they have been 
taken in CO, production. For instance, he found that one gram of 
Ricinus integuments exposed to light gave off 18 times more CO, 
than one gram of seeds with testas removed; that one gram of the 
testas of Vicia Faba gave off 10 times as much CO, as the same 
weight of decorticated seeds; and that one gram of the coats of 
the pea produced 25 times as much CO, as a gram of the embryos. 
