

1891.] The Comparative Morphology of the Fungi. 1061 
_ lies the true character of the chlamydospore, which has been lost 
in other groups, leaving it as the essential feature of the Uredi- 
nez and the related Ustilagineee. No species which has teleu- 
tospores can be considered “ incomplete,” the only such species 
being those whose teleutospores, and therefore whose basidia, are 
unknown. A single other fruit form is common among the 
Uredinez, the so-called “ spermogonium,” but its relations may 
better be discussed in connection with those of the similar struc- 
tures which occur abundantly among the Ascomycetes. 
The Ustilagineze have been a source of much perplexity as to 
their relationships, although the similarity of the spore-germina- 
tion in many forms to that of the Uredinee has indicated the 
propriety of placing these groups near together in the system. 
In their formation the spores of this group closely resemble those 
of Ptychogaster and other chlamydospores. In germination many 
of them produce structures which strikingly recall the basidia and 
spores of the Uredinez, and a few produce merely vegetative 
filaments, perhaps by degeneration from the former type; while 
those of a large group of forms give rise to undivided filaments, 
each with a whorl of conidia at its apex, representing clearly the 
basidia of the Autobasidiomycetes. In the great majority of these 
fungi the chlamydospores are the only fruit form developed, but 
some of them produce also conidiophores with typical. conidia. 
The basidia of the Ustilaginee are distinctly more primitive 
than those of any of the true Basidiomycetes, in that they are 
much more variable in form, size, and number of spores, in all 
those particulars,—that is, whose definiteness constitutes the true 
basidium. Brefeld therefore places the group, under the name of 
Hemibasidii, between the Phycomycetes and the true Basidiomy- 
cetes, as a connecting link; and divides it on the basis of the two 
types of basidia already described, into Protohemibasidii and 
Autohemibasidii, corresponding to the two groups of Basidiomy- 
cetes. These two groups coincide with the two families into 
which the smuts have been divided by Schroeter on the basis of 
Spore-germination, the Ustilaginei and Tilletiacei. 
Passing now to the other great group of fungi, we find it also 
€specially characterized by a particular form of reproductive 


