confluens and the Morphology of the Ascocarp. 389 
wall at its base. Such a change as this would be a very 
natural correlation of the multinucleated condition and con¬ 
sequent enlargement of the antheridial cell which we find in 
Pyronema. It is to be remembered that the long club-shaped 
antheridium of the latter is to be compared with the antheridial 
cell of Sphaerotheca , not with the entire male branch of the 
latter. It is easy to see that a cell containing the mass of 
nuclei present in the antheridium of Pyronema could hardly 
be perched upon the apex of the oogonium as is the small 
antheridial cell of Sphaerotheca. We can readily conceive, 
then, the development of the conjugating tube as a means 
of overcoming this difficulty. Still I am inclined to believe 
that in reality the reverse process has taken place, and that 
the sexual apparatus with the trichogyne represents the 
more primitive type for the Ascomycetes. 
Sphaerotheca and the Erysipheae, as I have already pointed 
out, may well be considered as a highly specialized group 
because of their parasitic habit and haustoria and their 
complex appendages, as compared with the relatively simple 
secondary mycelia of Ascobolns and Pyronema. 
In view of the difference between the sexual apparatus in the 
Erysipheae and that in Pyronema , it is very interesting to note 
that Ascobolns , in the structure of its fully developed asco- 
gonium, while superficially it is more like Sphaerotheca than 
Pyronema in one most important feature, is much more closely 
related to the latter. While in Sphaerotheca and Erysiphe 
only one cell of the ascogonium has its contents utilized in 
the formation of asci, the entire contents of the ascogonium 
of Ascobolns are emptied by means of the pores in its septa 
into the cell from which the ascogenous hyphae spring, and 
thus are utilized in ascus-formation. Similarly, as we have 
seen above, the entire contents of the ascogonium of Pyronema 
are utilized in ascus-formation. In view of these facts it 
would be extremely interesting to know exactly the nature 
of the earliest stages in the development of the ascogonium 
in Ascobolns. 
In the Laboulbeniaceae, as Thaxter shows (84), this differ- 
