104 
Journal of Agricultural Research 
Vol. XXI, No. 2 
In this series of samples there were included some seeds which had been 
killed by boiling, then dried, inoculated with spores, and stored in the same 
way. Some of these were broken in the usual manner over the embryo. 
The latter were attacked quickly, much more so than the living seeds 
similarly injured. The dead seeds with unbroken coats were attacked 
later, but considerably more quickly than the living seeds with perfect 
coats (Pi. 18, B). The order in which the mycelium became evident on 
these different lots of seed is as follows: 1 
1. Seeds killed by boiling, seed coats broken over embryo. 
2. Seeds living, seed coats broken over endosperm. 
3. Seeds living, seed coats broken over embryo. 
4. Seeds killed by boiling, seed coats unbroken. 
5. Seeds living, seed coats unbroken. 
The reason for the great susceptibility of wheat seed which has been 
scratched or broken over the endosperm is presumably clear. Rhizopus 
and Penicillium are saprophytes and grow readily in the exposed storage 
products of the endosperm. The embryo, however, is composed of living 
cells, and it is only under unusual conditions, if at all, that it can be 
directly attacked. Even in the few cases where it appeared to be directly 
infected, it is possible that the fungus reached the endosperm first and, 
developing there, killed the embryo by destroying its food or by poisoning 
it with the toxins produced. Entrance of the fungi into these seeds 
artificially injured over the endosperm results in death or injury to the 
embryo, probably by destroying the stored food rather than by actually 
entering the embryo itself. That this immunity of living embryos to 
direct attack is due to some property which disappears when death occurs 
is consistent with what we know of the behavior of plants in general 
toward saprophytes. 
The practical importance of these facts is that the loss of seed through 
attack by soil saprophytes is due largely to one or two causes, both pre¬ 
ventable: (1) The seed coats surrounding the endosperm are injured by 
improper handling (PI. 17, B), or (2) the embryo has been injured by 
seed treatment with copper sulphate or other fungicide. In the first 
case, the embryo or seedling is unable to develop, probably because of 
the destruction of its food supply in the endosperm, and in the second 
case, it is so weakened by the chemical that its natural immunity from 
saprophytes is lost. Extreme instances of the first condition have 
occurred frequently in California. Wheat has been sold for seed which has 
been subjected to a scouring process preparatory to milling and was so 
badly scratched that it would not germinate well, even if untreated, so 
long as there were saprophytic fungi to infect it (PI. 17). Such seed 
can not be germinated, even in sterile blotters, unless it is previously 
1 Seeds killed by boiling with seed coats broken over the endosperm were not included in these ex¬ 
periments because living seeds with endosperm injury are so very suscept ible to fungous attack that it 
was deemed superfluous to try the effect of killing such seeds before breaking the seed coats. 
