208 THE JOURNAL OF BOTANY 
of host-plants, and is composed of a number of ‘biol orms, 
we must investigate the pheno of specialized parasitism 
shown in the restricted infection powers the a ores an 
conidia of each “biologic form.” instance, in the case of the 
on Hordeum vulgare, the ascospores seem unable to infect wheat, 
oats, rye, Hordeum maritimum, H. secalinum, H. jubatum, and H. 
The possibility must not be lost sight of, however, that this 
“biologic form” may be able to infect certain other species of 
grasses, and from these pass on to wheat, oats, &e. When working 
with the Oidium of E. Graminis on species of Bromus, the results of 
certain experiments led me to believe that although certain Oidia 
may not be able to infect directly a host-plant, a, yet they may, be 
able to do so by means of their power to infect another plant, 4, 
from which the Cidium is able to infect a. Thus it was found that 
the “biologic forms” of Oidium on Bromus hordeaceus, B. interruptus, 
and B. commutatus are all incapable of infecting directly B. sterilis. 
They are all, however, capable of infecting B. tectorum. Now the 
, Or om fi 
B. sterilis, we shall have a clear case of a species of host-plant which 
es, from the point of view of the fungus, as a bridge between 
Whether in all cases the characteristics of “biologic forms” are 
igerous stages remains to be deter- 
mined by future experiments. In the present case the ‘ biologic 
of Bh. Gramini: ordeum vulgare has been found by 
Marchal to be characterized in its Oidium-stage by its power of 
infecting H. distichon, H. hexastichon, H. Zeocriton, H. trifur- 
catum, H. nudum, H, jubatum, and H. murinum, but not H. 
maritimum, H, secalinum, nor H., bulbosum, nor various other 
grasses belonging to twenty-seven different genera. In the experi- 
ments described above with the ascos 
it was found ar soeig i Zeocriton and H. trifurcatum were able ap 
in » Dut not H. maritimum, H. secalinum, nor H.bulbosum. § 
- far, then, the ch: 
