1HE RUSTS OF NOVA SCOTIA. FKASER. 323 



are not nearly related. Thus the species belonging to 

 Gymnosporangium have their aecial stage on a tribe of the 

 rose family while the telial stage is on conifers. The grass 

 and sedge rusts usually have their telial stage on the Com- 

 positae, though there are many exceptions. All the grass and 

 sedge rusts are heteroecious with one or two exceptions. Com- 

 mon autoecious rusts are those belonging to the genus 

 Phragmidium; other species are Puccinia Menthae, P. Violae, 

 Uromyces Limonii and U. Polygoni. 



When different forms occur together on the same host 

 plant they are often assumed to belong to the same species, but 

 it is not always safe to do so, without the test of infection 

 experiments. 



In the largest number of heteroecious rusts the teliospores 

 are formed at the close of the season or of the life of the host, 

 and rest during the winter. The aecial host is infected in the 

 spring, and the alternate host from the aeciospores. That is 

 the case in Puccinia graminis and many other rusts. In other 

 eafc-es the telial .mycelium hibernates, the teliospores are pro- 

 duced in the spring and germinate immediately, and the 

 basidiospores infect the aecial host, as in Chrysomyxa ledicola. 

 Again the telial mycelium may be actually perennial, the 

 teliospores formed in the spring and germinating infect the 

 aecial host, the aecial mycelium lasting only during tJie sum- 

 mer as in Gymnosporangium. There are other types in which 

 the aecial mycelium hibernates, or the mycelium of both 

 generations may be perennial. 



The aecia of heteroecious rusts develop only from mycelium 

 formed by basidiospore infection, the aecia do not repeat, nor 

 can the basidiospores produce infection in the host plants in 

 which the telia are produced. The mycelium that bears telia 

 and uredinia arises only from infection by aeciospores or 

 urediniospores. The rust, however,- may be propagated by the 

 urediniospores for some time, probably indefinitely. In Aus- 

 tralia P. grarninis has lost the power to infect the barberry 



