482 Journal of Agricultural Research voi. x. No. 9 



Hosts on which P. graminis phleipratensis was found in nature: Dac- 

 tylis glomerata L., Festuca elatior L., F. pratensis Huds., Koeleria cristata 

 (I/.) Pers., Phleum pratense ly. 



Hosts heavily infected by artificial inoculation: Alopecurus genicula- 

 tus h., A. pratensis L., Holcus lanatus L. 



Weakly infected by artificial inoculation: Avena sativa h-, A. fatua 

 L., Arrhenatherum elatius (L.) Beauv., Bromus tectorum L., Elymus vir- 

 ginicus h., Hordewm jubatum h., H. vulgar e L., Lolium italicum R. Br., 

 L. perenne L., Secale cereale L. 



Inoculated but not infected: Agropyron repens (L.) Beauv., Agrostis 

 alba ly., Festuca ovina L/., F. rubra L., Poa compressa L., Triiicum vul- 

 gar e Vill. 



It will be observed that the timothy rust occurs in the field on Dac- 

 tylis glomerata, Festuca elatior, and F. pratensis and probably Koeleria 

 cristata, in addition to Phleum pratense. Eriksson and Henning (13, p. 

 130) give only Phleum pratense as a host plant, as does Johnson (19), 

 although he succeeded in infecting D. glomerata and F. elatior by means 

 of artificial inoculations. How commonly this rust occurs on D. glomerata 

 the writers do not know; however, it does occur commonly on both 

 species of Festuca mentioned. 



The character of infection on the plants mentioned as weakly infected 

 in the greenhouse has been previously discussed (26) and need not be 

 repeated here. 



The relation of the rust to P. graminis is still probably debatable. The 

 problem has been summarized by Stakman and Jensen (26, p. 211). So 

 far as the writers have been able to determine, no one has yet been able 

 to infect barberry consistently with the teliospores. The writers made 

 numerous inoculations in the spring and early summer of 191 6 at a time 

 when teliospores of the common biologic forms readily infected barber- 

 ries, but no secia developed as a result of the inoculations made with 

 timothy-rust teliospores. Flecks were commonly developed, and, al- 

 though histological examination was not made, it is quite probable that 

 infection occurred, but the rust was unable to develop secia. On mor- 

 phological grounds there seems to be no justification for regarding 

 timothy rust as a distinct species ; it does not differ more from ordinary 

 biologic forms of P. graminis than does P. graminis agrostis. The same 

 is true of its infection capabilities. The writers favor including it as a 

 biologic form of P. graminis. 



The distribution of the rust is interesting. In 191 1 Johnson (19, p. 7) 

 stated that it was common at least as far west as Minnesota and in 1914 

 Mercer (21, p. 20-22) recorded it as being common in North Dakota. 

 In 1 91 6 the writers found it very commonly in serious abundance in 

 Minnesota, North Dakota, Montana, Idaho, Washington, Wyoming, 

 Nebraska, Iowa, and Manitoba, Canada. 



