Heed—Infection Experiments with Mildew. 529 
ment. According to my observations, this is not sufficient to 
prove that these hosts are immune to the fungus. 
Salmon (28, 30) has also obtained some interesting results 
by injuring host plants in various ways and then inoculating 
them with conidia which would not infect them when unin¬ 
jured. He has found that if barley leaves are mechanically in¬ 
jured in any way, or if treated with ether, chloroform or 
alcohol, or if heated in water to about 50 degrees C., and then 
inoculated with conidia from wheat ( Triticum vulgare ), they 
lose their power of resistance to the wheat mildew and become 
infected, although healthy normal leaves of barley cannot be 
infected. However, the conidia produced un the barley leaves 
as a result of these methods of treatment are not capable of 
causing infection on uninjured barley leaves, but they do in¬ 
fect normal wheat leaves, the host from which the conidia 
were first taken. 
It has also been found by Salmon (32) that the mildew on 
wheat will infect young leaves of Hordeum sylvaticum. He 
has cultivated the mildew on H. sylvaticum for five genera¬ 
tions, but the conidia produced on the new host never lost the 
power to infect the wheat from which the fungus was origin¬ 
ally taken. Furthermore, it did not acquire the power to in¬ 
fect the hosts, such as barley, which the mildew found in na¬ 
ture on H. sylvaticum possesses. We thus see that the fungus 
is not changed in any way as a result of living on this unusual 
host. 
It has further been shown that a fungus can get a start on 
a host plant although it may not be able to cause complete in¬ 
fection. Salmon (33) finds that if the mildew from wheat, 
for example, is sown upon barley, the fungus may penetrate 
the cells of the barley and in some cases form fully developed 
haustoria. More commonly the haustoria' are arrested in their 
development, remaining as more or less rounded refractive 
bodies in the cells, which soon completely disorganize. Salmon 
concludes from these facts that the susceptibility or immunity 
of a host plant does not depend on the presence or the absence 
of a chemotactic substance in the cells of the host, but “on 
34—S. Sc A. 
