Some Suggestions as to the Phylogeny of the Ascomycetes 
151 
of tissue in which the carpospores are borne shows a remarkable simi¬ 
larity to the apothecium of some Lichen I do not want to be understood 
as advocating any claim for Harveyella in the ancestral line of the Asco¬ 
mycetes but I wish to show that the modern Florideae are capable of 
giving rise to parasites with many of the fungal characteristics. If this 
is possible with recent Florideae it does not seem reasonable to deny the 
possibility of the more primitive Florideae having given rise to a hystero- 
phytic offshoot that eventually, became the Ascomycetes. 
The most difficult objection to overcome is the ascus. No living 
Red seaweed is known in which such a structure is present. True, it 
has been suggested that the tetraspore mother cell represents an ascus 
and the tetraspores four ascospores. An additional point of similarity is 
the fact that in both structures, viz. tetrasporangium and ascus we find 
the occurrence of a reduction division. However, in view of the association 
of the tetrasporangium with a type of alternation of generations such as 
appears not to occur among the fungi except in the Uredinales, and of 
the totally different origin of tetrasporangium and ascus morphologically, 
I feel that we -must for the present look elsewhere, unless we are to 
suppose that the ancestors of the Ascomycetes suffered a shortening of 
the tetrasporic generation and combination of it with the sporophytic 
structure arising from the fertilized carpogone, so that the sporogenous 
filaments from the carpogone, instead of terminating in carpospores came 
to terminate in tetrasporangia (asci). Perhaps Nemalion which does not 
possess an intercalated tetrasporic generation may give us a hint as to 
the possible origin of the ascus. In this form the fusion nuclei formed 
by the division of the fertilized nucleus of the carpogone are diploid in 
character. The rather short sporogenous threads are also composed of 
cells with diploid nuclei. Wolfe 1 ) has found, however, that the end cell 
of each sporogenous filament, the cell from which the carpospores are 
budded off, shows the haploid number of chromosomes in the divisions 
which give rise to the carpospores. He found no indication of tetrad 
formation or of the two nuclear divisions always considered as essential 
to the reduction of chromosomes. Possibly the carpospores thus produced 
contain the haploid number of double chromosomes and on germination 
make the typical reduction divisions or possibly these take place just back 
of the cells on question. In either event those cells in which this occurs 
show considerable homology to the asci which are cells produced near 
the ends of filamentous outgrowths from the fertilized carpogone, within 
which the reduction division takes place. Perhaps the retention and 
further division of the products of this division within the ascus so as to 
form eight or even more ascospores is connected in some way with the 
changed manner of spore distribution. In the seaweeds the comparatively 
large carpospores are easily transported by the water while the air- 
dispersal is more efficacious with smaller spores, such as the ascospores. 
In the Florideae the union of the nuclei takes place in the carpo¬ 
gone. The sporogenous threads, then, contain daughter fusion nuclei. In 
some of the Ascomycetes doubt has been thrown recently upon the occur¬ 
rence of this union within the carpogone. Indeed Claussen, Schikorra 
1) Wolfe, J. J., Cytological Studies on Nemalion 
18 [Oct.], 607—630; 1 textfig., pi. 40—41). 
(Ann. of Bot. 1904, 
