GUASSE^ OF IOWA. 269 



bearing lateral spores, sporidii. These sporidia, when in con- 

 tact with the barberry leaf, enter by boring their way through 

 the epidermal cells. 



The barherry cluster cup fungus, anclits connection with common 

 grass rust. — It is not absolutely necessary for the common grass 

 rust to have its first stage on the barberry, yet experiment has 

 shown beyond doubt that it does occur on that plant. The 

 theory has been advanced that appearing in one of its stages 

 oa the barberry gives the parasite new vigor. It is not improb- 

 able that in some places the mycelium or vegetative part of 

 the fungus may be perennial in the tissues of grasses, as it is 

 with ma -y other fungi, and I am inclined to think this is true in 

 southern localities. Beyond question this rust produces spores 

 during the entire year in our southern states, and on the 

 approach of early spring gradually moves northward. I may 

 also mention the facb that this rust certainly does not in the 

 west appear before the cluster cup fungus on the barberry 

 appears. It is usually eight or ten days later, and then only 

 to a limited extent. Rust often appears where barberry does 

 not occur within hundreds of miles. This was especially 

 noticeable during the early history of grain culture in the 

 northwest. Rust fol.ows a general infection. 



Grasses affected. — In addition to the cereal plants enumerated 

 in the table, this rust has been reported on the following 

 grasses: Bent grass, Agrostis alba, A. capillaris, A. canina, 

 Aira caespitosa, Alopecurus pratensis, Agropyron rejiens, A. spica- 

 tum, A. caninum, Avena fatua, Briza media, Bromus tectorum, Briza 

 maxiina, Bromus mollis, Galamagrostis, Dactylis glomerata, Dis- 

 tichilis maritima, and Eriksson and HenniDg report it also on 

 Elymus canadensis and several other forms; it has, however, not 

 been seen by me on this host. We have repeatedly seen it on 

 Holcus lanatus and Hordeum jubaturit.* 



It should be borne in mind that in many cases these hosts 

 have not been determined by inoculation experiments. The 

 inoculation experiments of Carleton show that the P. graminis 

 avence are not successful when applied to wheat, but successful 

 when applied to Avena sativa, A. fatua, A. pratensis., Dactylis 

 glomerata, Arrhenatherum avenaceum. The following are probable 

 hosts: Koeleria cristata, Ammophila arenaria, Bromus ciliatus 

 and Loliam perenne. The Hordeum jubatum, according to Carle- 

 ton, supports two distinct forms of P. graminis. The same 



*auU. Iowa Agrl. Exp, Sta. 30: 302. 1895. 



