1909] STEVENS & HALL—VARIATION OF FUNGI 19 
since flooding the plate with an extra tube of agar after the agar first 
plated had set, did not change the proportion of the two types. Nor 
did sowing in such way that the spores were at the bottom rather than 
at the top of the agar change results. There was a marked tendency 
of colonies of both types of the fungus to become more productive of 
large pycnidia where two different colonies approach each other, sug- 
Fic. 14.—A — Chrysanthemi Stevens; portion of colony showing many 
pycnidia; cf. fig. 
gesting that = might be needed a cooperation of two diverse 
strains in order to form a pycnidium; that the strains of few pycnidia 
lacked the requisite individuals, and that the strains of many pycnidia 
had more than one individual to the colony. To test this, colonies 
were traced from the earliest development, resulting in clear evidence 
that in some instances a colony developed from a single spore was one 
with few pycnidia; in other instances a single spore produced a colony 
of many pycnidia. 
CoNIOTHYRIUM FUCKELIT SACC., FROM APPLE 
In one instance this fungus, which rarely fruited, made pycnidia in 
almost perfect circles near the margins of each colony on the plate 
(fig. 15). 
