48 CEREAL RUSTS OF THE UNITED STATES. 



which, having" less resistant constitutions, are always more susceptible 

 to disease than older ones; and (2) as the experiments were made in 

 the greenhouse the conditions were, of course, somewhat more favorable 

 to rust propagation than those outdoors. As to the condition of the 

 host, the matter was tested in a few cases by inoculating at a more 

 advanced age certain x^lants that had been successfully infected when 

 quite young, the result being that infections did not occur at all or were 

 very slight and evidently took place with difficultj^ In the experiments 

 with the leaf rusts of wheat and rye young plants were also used as 

 hosts, but the results were quite different. The crown rust, it seems, 

 is not yet sharply limited in its adaptability^ to particular hosts. 



The table shows that although rust from oats infected a number of 

 wild grasses, no case is recorded of an infection of oats from a wild 

 grass; hence it has not been demonstrated that the crown rust of oats 

 is equivalent to the rust of any wild grass, although it is of course 

 quite probable that the same rust occurs on several species of the genus 

 Avena and perhaps on other grasses. In the writer's opinion the iden- 

 tity of a rust of anj^ grass with that occurring on any cereal is not 

 established until reverse infections have been produced. 



Another important question that yet remains unsettled is whether 

 there is an ;ecidial host for this rust of oats in the United States. 

 True, an aecidium on a species of Rhamnus has been collected in a 

 number of places, but as yet no inoculation experiments have been 

 made with it, and it will require many to determine accurately the 

 status of the crown rust in this country.^ 



The usual incubation period to produce infections on oats in case of 

 inoculations with this rust is from seven to ten days, but to infect cer- 

 tain other plants, such as Hordeum murinum and BactyJis (/Jomerata, a 

 longer time is required, and when plants are rather old scarcely any 

 Infection results. 



From experiments so far made, only cultivated varieties of the fol- 

 lowing subspecies oi Avena sativa can be considered to be hosts of this 

 rust in the United States: Avena sativa patula, A. sativa orientaUs, and 

 A. sativa nuda.- Other species of Avena that will piobably j'et be 

 found to act as hosts are A.fatua, A. pratensis, A. hooTceri, and A. 

 sterUis. Of course there is the possibility also that this rust will be 

 found to have several specialized forms in this country, as Eriksson 

 (28, p. 302) has already found to be the case in Sweden. 



Occurrence and distribution. — The crown rust in its distribution seems 

 to be quite analogous to the leaf rust of wheat, and in that regard both 

 bear somewhat similar relations to the black stem rusts of the same 



I Since this bulletin was iirepared the writer has made a number of inoculation 

 experiments with the iecidiuni of I'hamnus Janceolata at Lincoln, Xebr., which resulted 

 in the infection of oats, Phalaris caroliniana, and Arrhouithcrum elaiius. 



"^Fhalaris caroliniana and Arrhcnaihernm (latins must now also be included as 

 probable hosts (see preceding footnote). 



