2l8 
AMERICAN JOURNAL OF BOTANY 
[VoL 9 
The last two sowings, of August 5 and August 25, never reached the heading- 
out stage. The barleys were rust-free until the middle of September, 
when a few uredo pustules were to be discovered on the leaves and sheaths 
of the three older plots. No teleuto was found. 
As has been noted for the plots of wheat and rye, even more strikingly 
in the case of the oats, the rate of development of the parasite differs with 
the host. It is much more rapid in the case of the oats than in that of the 
other cereals. For example, in the seedling stage the amount of rust on 
the plants may appear less on the oats than on the rye and wheat, although 
at the time of heading out the same plants will show the reverse condition, 
the oats being much more severely infected. 
The rust history of plots of cereals sown at successive intervals through 
the summer may be taken as indicating that the age and maturity of the 
host is a factor in the progress of the disease, and that the action of this 
factor differs with the identity of the host plant. 
Laboratory and Greenhouse Studies 
Culture Methods 
Four cereal rusts — Puccinia coronifera Kleb., P. secalina Grove, P. triti- 
cina Eriks., and P. Sorghi Schw. — were successfully grown for periods of 
time on the host in pots on greenhouse benches as described by Melhus 
(1912) and Fromme (1913), and under aseptic conditions on host seedlings 
growing in test tubes as described by Ward (1902a) and Mains (1917). 
Variations were introduced in both methods. 
Fromme (191 3) reviews the problem of growing cereal rusts in the green- 
house. The method recommended by him includes sowing rust spores on 
new host plants every few weeks by applying them with a scalpel or camel's 
hair brush, or spraying on in suspension in water with an atomizer, and 
then putting the host plants into a moist chamber for from 24 to 48 hours 
to provide the conditions of high humidity necessary for spore germination 
and infection. Tests, however, indicated that the first part of the method 
recommended by Fromme, artificially sowing rust spores on the new host, 
was not necessary under the conditions obtaining in the Columbia green- 
house. It was found that when new host plants are grown beside infected 
plants in the greenhouse, rust spores will be sown on them by natural 
agencies, such as convection and other atmospheric currents, sufficient to 
produce abundant infection if conditions of high humidity are provided 
occasionally to render possible the germination of the spores. 
Accordingly, the method adopted for maintaining stock cultures of the 
cereal rusts in the greenhouse was to introduce new host plants alongside 
the infected plants every third week and to cover the cultures with a moist 
chamber every second or third night. The fungus maintained itself self- 
sown in this manner, and no artificial inoculations were needed. The 
