86 
Journal of Agricultural Research 
Vol. XXVIII, No. 1 
Table VII shows a reduction of 31.7 per cent in the germination of the treated 
machine-threshed seed, while the treated hand-threshed seed was reduced only 
1.1 per cent. This difference of 30.6 per cent between the germination of machine- 
threshed and hand-threshed seed shows clearly the relation of mechanical injury 
of the seed coat and possible subsequent injury from application of hot water 
(PI. 4). 
Having shown before that unbroken seed coats protect the grain from treat¬ 
ment injury, it may be asked why treated hand-threshed wheat (Table VII) did not 
germinate fully as well as untreated hand-threshed seed. It is possible that some 
injury to seed coats may result even from hand threshing. Also, localized im¬ 
perfections too slight even to be seen microscopically, may occur in seed coats. 
Shull {12) found such a condition to be frequent in the seed coats of Xanthium. 
Hurd (5) observed that wheat kernels with apparently unbroken seed coats 
occasionally were injured when treated with copper sulphate. 
In an experiment to determine if a break in the seed coat invariably could be 
detected microscopically and thus enable the observer to predict the amount of 
injury likely to result from treatment, a machine-threshed sample of Haynes 
*Bluestem (C. I. 2874) was employed. The untreated seed germinated 95 per cent. 
One hundred kernels selected at random and treated, germinated 12 per cent, 
while a like number of kernels microscopically examined, selected for unbroken 
seed coats, and treated simultaneously with the above, germinated 74 per cent. 
This indicates that it is hardly possible to detect all of the breaks in a seed coat 
even under the microscope, and that the small fractures which escape detection 
may result fatally to the embryo, through injury from treatment. Wallden {15) 
was able to detect broken seed coats by the use of eosin, which does not penetrate 
^he unbroken seed coat, but enters every small fissure, staining the seed tissues 
at the point of entrance. 
THE NATURE OF SEED-COAT PROTECTION FROM HOT-WATER INJURY 
The important r61e of the seed coat in protecting the embryo from injury dur¬ 
ing the hot-water treatment led to an investigation of the manner in which it 
functions. Obviously, this protective action must be exerted either in the 4 to 
6-hour presoak in cold water, in the 10-minute bath in water at 54° C., or in both. 
In an experiment to determine the point, three lots of 50 hand-threshed kernels 
of each of the three varieties, Leap (C. I. 4823), Poole (C. I. 3489), and Power 
Fife (C. I. 3697) were used. In the first lot the seed coats were left unbroken 
through both the presoak and 10-minute baths. In the second lot the seed coats 
were broken over the embryo before the 10-minute bath. In the third lot the seed 
coats were broken over the embryo before both baths. The presoak bath for 5 
hours at 20° C. was applied to all lots simultaneously on February 10, 1923, and 
each lot was weighed carefully before and after presoaking to determine the pos¬ 
sible influence of broken seed coats on the amount of water absorbed by the seed. 
All lots were treated then simultaneously for 10 minutes at 54° C., dried 6 days at 
room temperature, and then sown on February 16, 1923, in soil in greenhouse 
flats. One month later, on March 16, the seedlings were counted. The results 
are presented in Table VIII. 
