112 
Journal of Agricultural Research 
Vol. XXI, No. a 
the embryo before it can be neutralized. Slightly damaged seed coats 
delay the entrance of the copper sulphate enough so that the lime saves 
the wheat. 
There was more injury to the plumule when the seed was injured over 
the plumule end of the embryo than when it was injured over the radicle. 
The explanation is probably to be found in the fact that copper sulphate 
may not reach the inner end of the embryo at all, because of a bubble of 
air in the little pocket where the plumule lies which might protect it 
from being wetted by the liquid. When the seed was injured over the 
plumule, the copper sulphate had easier access to it and produced ex¬ 
treme deformity. When harvester-thrashed wheat is treated, the injury 
appears in the stunted roots of the seedling. More rarely is there a 
shortened or distorted plumule (PI. 23). 
An examination of several lots of harvester-thrashed seed shows great 
variation in the degree to which the seed coat over the embryo is injured. 
The question naturally arises as to whether the slight cracks which are 
barely visible allow a copper-sulphate solution to penetrate or whether 
the inner layer of the seed coat is a semipermeable membrane and keeps 
the copper sulphate out, even though the outer one is ruptured. A study 
of the seed coat over the embryo shows it to be composed of two layers 
which can be split apart and torn off separately. Bollev (2) gives a figure 
of a longitudinal section through a wheat kernel illustrating the difference 
in the structure of the coat over thejembryo and the endosperm. To de¬ 
termine the semipermeability of the inner of these two layers, the outer 
was slit with a sharp needle so as to make a distinct tear, taking care 
not to injure the inner layer. Twenty-five such seeds were then dipped 
in a 1 to 4 copper-sulphate solution for 4 minutes and drained for 15 
minutes. Another lot was broken in the same way but through both 
layers of the seed coat, and a control of uninjured seed was used 
(PI. 22, A, 4). The germination percentages foliow: 
Per cent. 
Both layers of the seed coat over the embryo broken. 30 
Only the outer layer of the seed coat broken. g Q 
Seed coat uninjured. IOO 
It appears from this experiment that the inner layer is impermeable to 
copper sulphate and that little or no injury results to the seed in treating 
unless both layers are broken. The slightly lowered germination of the 
less-injured seeds no doubt is due to accidental injury to the inner layer. 
Thus is explained the fact, that the percentage of mechanically injured 
seed in a harvester-thrashed lot is slightly higher than the percentage of 
such seed killed by treatment. The percentage of badly injured seed 
counted corresponds more nearly to the amount of injury to be expected 
when the sample is treated with a 1 to 4 solution. 
It is interesting to compare the percentage of injured seed coats found 
ih some of the samples of harvester-thrashed seed sent to the laboratory 
