228 BOTANICAL GAZETTE [SEPTEMBER 
leaved pink which was known to be very susceptible to the carnation rust, 
was selected. Sets of cuttings were taken from the stock plant and rooted 
in river sand. After the cuttings were well rooted, they were transplanted 
to pots containing the soils to be tested. Each of the soils was carefully 
mixed before it was put into the pots, so that all would be as uniform in 
composition as possible. The five soils used ranged from one that was - 
principally sand with a very small amount of organic matter to those 
containing less sand and more organic matter and clav. 
._ After the plants were potted, they were arranged im sets of five or seven 
in large saucers, and placed so far as it was possible under the same condi- 
tions of light, heat, etc. When the plants had become established and 
made a growth of a few inches, each set was inoculated. Of the 170 plants 
inoculated, only three failed to show rust sori in sixteen to twenty-one days, 
the raajority showing sori in seventeen to nineteen days. 
The leaves were carefully examined with a hand lens twice a day after 
yellow spots began to show on them, indicating that infection had taken 
place. When the uredospores were observed to be breaking through the 
epidermis of a leaf, a record was entered for that plant. This record was 
afterward used in making comparisons with the composition of the soils. 
The Bureau of Soils of the United States Department of Agriculture made 
a mechanical analysis of the soils and determined the percentage of organic 
matter and nitrogen. 
Rather than depend upon his own judgment, which was liable to have 
been somewhat biased on account of having watched the development of 
the pinks and the rust, the writer averaged those of several other persons 
with respect to vigor, growth, and color of the plants. 
The results of the investigation are summarized in part below; the 
details will be published later. Whether similar results can be obtained 
by using a different host and parasite remains to be determined. 
1. The intensity of color was directly proportional to the amount cf 
clay in the different soils. 
2. The growth of the host was directly proportional to the amount of 
organic matter, nitrogen, and silt in the different soils. : 
: 3- The period of incubation of the carnation rust, while not uniform 
mM every instance, was in general inversely proportional to the amount 
of organic matter, nitrogen, and silt in the different soils, and to the growth 
of the host; it was directly proportional to the amount of gravel and sand ” 
the different soils; that is, the more gravel and sand there were in a 
the longer it was before the uredospores broke through the epidermis after 
an inoculation had been made; and the more organic matter, nitrogen, and 
