42 PUCCINIA GRAMINIS 
culms; for these differences the systematic part can be con- 
sulted (Fig. 25). 
There is, however, one point 
of difference connected with P. 
graminis which possesses great 
biological interest—its virtual in- 
dependence of the ecidial stage. 
LJ For a long time it had been 
known that Barberry bushes in 
the hedges caused “mildew” on 
the corn in the neighbouring 
c fields, and when, in 1864-5, De 
\ 
Bary proved the hetercecism by 
i | experimental cultures, it was too 
Big. 25. Puccinia graminis. a, hastily assumed that the ecidium 
ecidia on Berberis; b, uredo- on the Berberis was just as es- 
spore ; c, teleutospores. : 
sential to the rust on the corn as 
that on the Nettle is to the rust on the Sedge. Many facts 
now tend to show that this is not the case. 
In Australia and the plains of India the Barberry is un- 
known except as an introduced plant, yet the Puccinia occurs 
everywhere and does enormous damage. McAlpine records, in 
his Rusts of Australia that he made numerous attempts to 
infect imported species of Berberis with the rust of Australian 
wheat which is morphologically undistinguishable from the 
P. graminis of Europe, but all his efforts were in vain. The 
inevitable inference is that P. graminis, as it occurs in those 
countries, is a “ biological” race which maintains itself by other 
than the primitive means. A similar thing is true, according 
to Lagerheim, in Ecuador, where also rust flourishes and does 
great harm. ; 
The facts now known concerning the specialisation of the 
Black Rust are treated of in a separate chapter, but there is 
one point which must be mentioned here. This concerns the 
mode by which fresh epidemics are produced each year. Even 
if the Barberry is present, it is by no means certain that it 
plays any important part in these annual attacks. Apart from 
that, there are several possibilities: (1) the fungus may winter 
ore ‘bd 
3. a 
