BIOLOGIC SPECIALIZATION IN THE GENUS SEPTORIA 
21 
mostly along the leaf margins, but often the whole leaf is involved. In the 
greenhouse growth is similar, but more rapid and extensive. Inoculated 
leaves are generally completely destroyed. The fungus is also reported 
upon Cirsium discolor. 
According to the infection experiments conducted, Septoria cirsii can 
make a weak attack upon both Cirsium discolor and C. lanceolatum. The 
spots formed upon these hosts resemble those upon C. arvense, but are 
smaller and very few in number. The difference in the vigor of attack 
upon these hosts is also well demonstrated by the length of the incubation 
period and the rate of development. Upon Cirsium arvense spots appeared 
in thirteen days after inoculation; pycnidia were observed in fifteen days, 
and in four weeks the whole plant was dead from the spread of the disease. 
Upon Cirsium lanceolatum spots were first visible in twenty days, and in 
four weeks they had attained a breadth ranging from 3-8 mm. Pycnidia 
and spores were found at this time. The plant had to be kept in a moist 
atmosphere to bring about the formation of many spores. The incubation 
period and rate of development were similar in Cirsium discolor. There 
were no noticeable difi^erences in the morphology of the fungus upon the 
various hosts. 
The relation of maturity of the spores to the presence of septa was often 
noticed in this species. Where spores were made to develop abundantly 
in a moist chamber, septa could not be seen at all, or could be seen only 
Diagram 9. Infections with Septoria Diagram 10. Infections with. Septoria brunellae 
cirsii from Cirsium arvense. from Prunella vulgaris. 
faintly after staining with iodine. In old spots, or where the growth had 
been slow, the septa were plainly seen. Spores obtained from an old culture 
upon corn meal agar had septa that were strikingly definite. Plants inocu- 
lated with spores from this culture were heavily infected. 
Septoria brunellae Ell. & Halst. 
The inoculations with Septoria brunellae gave negative results in all 
cases except upon the original host, which facts indicate that the fungus is 
probably limited to Prunella vulgaris in its host range. Upon this one host 
