486 Journal of Agricultural Research voi. x, N0.9 



while those of the remaining two forms are very much smaller. The 

 spores of the avenae and secalis forms do not differ greatly in length, but 

 the spores of the form avenae are considerably thicker than those of the 

 form secalis. The spores of the form agrostis are smaller than those of 

 the form phleipratensis, especially in width. 



APPEARANCE AND DEVELOPMENT OP UREDINIA 



The general appearance of the uredinia of different biologic forms of 

 stemrust on congenial hosts is, with minor exceptions, quite similar. The 

 linear uredinia of P. graminis phleipratensis are, of course, distinctive, 

 but even this character is not constant. The shape of uredinia on con- 

 genial hosts is usually oval or somewhat broadly linear, although, when 

 heavy infection occurs, the uredinia may coalesce to such an extent that 

 the identity of the individual uredinia is almost lost. The length varies 

 from only i or 2 mm. to 0.5 cm. or more. The color is quite variable. 

 Usually when the relative humidity is high or the light intensity low the 

 color tends to become paler. On some hosts the epiderm is raised be- 

 fore the uredinia break out, giving the plant a blistered appearance in the 

 infected area. The torn edges may persist for some time after the epi- 

 derm has been ruptured. Again, this appearance may be entirely absent. 



The size, shape, and color of uredinia may vary greatly on semi- 

 congenial or uncongenial hosts. Very often the uredinia become rounder 

 and very much smaller on such hosts. The color may also vary a great 

 deal, so much, in fact, that it is sometimes necessary to transfer the rust 

 back to the original host in order to become convinced that something 

 has not gone wrong. Carleton (7, p. 16) long ago called attention to this 

 fact, and it has been a matter of common observation since then. Of 

 course, such differences are to be expected, just as one expects phanero- 

 gamic plants to vary v.dth the conditions under which they are grown; 

 but with the rusts this variation emphasizes the necessity of determining 

 the performance before the identity can be established. 



The incubation period of the various forms varies considerably on the 

 same host. The age of the plant, the vegetative condition of the plant, 

 the temperature, and the light intensity may affect the rate of develop- 

 ment of the rust materially. The cereals are usually susceptible at any 

 age up to ripening time. The susceptibility of some grasses, however, 

 under greenhouse conditions at least, seems to vary more. No careful 

 experiments were made to investigate this question thoroughly, but 

 repeated observations Vv^ere made under conditions which made accurate 

 comparisons possible. Some plants, such as some species of Agropyron 

 and Elymus, are extremely susceptible when young and much less so 

 when older. In fact no infection, or at best very weak infection, occurred 

 on some of these grasses when they were about 3 months old; while. 



