AMg. 27, 1917 Piiccinia graniinis on Cereals and Grasses 489 



made by the writers at St. Paul, Minn., in the winter of 1916-17, an 

 unusually severe one, indicate quite clearly that the urediniospores sur- 

 vive the winter quite easily in Minnesota. A summary of the experi- 

 ments made on the overwintering of various cereal rusts is given by 

 Freeman and Johnson (14, p. 45-53), who also found that P. graminis 

 and probably P. rubigo-vera could survive the winter in the uredinial 

 stage at St. Paul, Minn. Bolley and Pritchard (4, p. 642) had previously 

 shown this to be true for P. rubigo-vera in a number of regions. From 

 preliminary observations it seems quite probable that urediniospores of 

 P. graminis are injured by heat and drying more than by low tempera- 

 tures. If this is true, the extent to which the rust overwinters may 

 depend on the amount of rain in the fall and on the amount of snow in 

 the winter. It would be very interesting to make extensive observations 

 in regions of high altitude, where snow comes early. Abundant uredi- 

 niospores occur on various grasses in the fall when the snows begin, and 

 it is possible that in such regions overwintering is very common. Further 

 investigations are needed. 



There can be little question that grasses aid very materially in the 

 dissemination of rust. Enormous numbers of urediniospores are devel- 

 oped on grasses, and these are undoubtedly blown long distances by the 

 wind. "Whether the grasses are important in initiating the early spring 

 infections, however, is another matter. If the uredinial stage of the 

 rust persists on grasses throughout the winter, they probably are responsi- 

 ble for much of the early infection. Even if such overwintering is of 

 infrequent occurrence, however, grasses may still be important because 

 they usually are nearer barberries than grain fields are. They may 

 become infected, develop urediniospores in a week or 10 days, and these 

 may then blow to other grasses or cereals and thus spread during the 

 entire growing season. 



An excellent summary of observations on the effect of barberry on 

 surrounding cereals and grasses is given by Freeman and Johnson (14, 

 p. 28-45) and by Pritchard (22, p. 169-175) and need not be repeated 

 here. Pritchard (22, p. 179-181), as a result of experiments in North 

 Dakota, concludes that barberry plants have but little effect in starting 

 rust epidemics in the spring, because — 



P. graminis probably appeared upon the experimental plot of winter wheat almost or 

 quite as early as upon Agropyron repens and Hordeum jubatum even when the latter 

 were in the immediate vicinity of the barberry. * * * With the exception of the 

 one case mentioned under date of June 2 7 , the uredospores of P. graminis were gen- 

 erally present upon the spring wheat earlier than they were observed upon the wild 

 grasses remote from the barberry bushes. 



He states further — 



that one form of P. graminis is common to Hordeum jubatum, Agropyron tenerum, 

 A. repens, Avena fatua, oats, and rye, but is incapable of infecting either barley or 

 wheat. This fiunishes little encoiuragement to those who believe that P. graminis 



