. 295 
reliable, i. e., a Sphzerotheca having two asci in a perithecium or an iry- 
siphe lamprocarpa having more than two spores in an ascus will not har- 
monize with a classification denying such variableness. Hence revisions 
are constantly being made, which have new characteristics as bases for 
new classifications, or extend the limits of these formerly too restricted 
species. But whatever advances may be made, the revisions must retain 
much of the old in the development of the new, for the first classification 
was correct in the main, and can be altered only in respect to details 
based on the more minute structure revealed by further investigation with 
greatly improved apparatus. 
