170 DopGE: RELATIONSHIPS’ OF FLORIDEAE AND ASCOMYCETES 
So far the crosier has not been homologized with any structure 
in the life history of the red algae. It seems to me that the pos- 
sible relations of its characteristic features and the resulting 
fusion of non-sister nuclei to the secondary fertilizations of Schmitz 
deserve careful consideration as I shall point out in more detail 
later. 
If one assumes that the outgrowth from the odgonium in 
Sphaerotheca represents simply a large, well-nourished ascogenous 
hypha, then a perfect type of crosier occurs here also. ‘The ascus, 
according to Harper (52), is formed from the binucleated penulti- 
mate cell of this outgrowth (Fic. 8, A). Uninucleated ultimate 
and antipenultimate cells are regularly formed. 
A few forms have been reported in which it is claimed that the 
ascus arises directly from the end cell of the ascogenous hypha 
without any bending over of the terminal portion. Maire (60) 
describes such a process in Galactinia succosa. The asci of 
Penicillium are described by Brefeld (16) as arising in rows from 
cells of the ascogenous hyphae. The same is true for Exoascus 
(Dangeard, 31). Thaxter (88) and Faull (39) have described the 
formation of the asci from the egg in several species of the Laboul- 
beniaceae. Here the ascogenous cell buds out directly into a 
large number of asci. Faull suggests several possible homologies 
for the cells developed from the germination of the egg, but he by 
no means makes clear their relationships to the parallel stages 
in the red algae. 
Ramlow (73) points out that the ascus in Thelebolus arises 
directly from a large binucleated cell of the ascogonium the other 
cells of which are uninucleated. In some cases the ascogenous 
cell is the subterminal cell. 
H. B. Brown (18) has recently reported that the asci of Xylaria 
are simply outgrowths from the end cells of the much-branched 
ascogenous hyphae. His description is however very brief and 
should perhaps not be quoted as proof of a departure from the 
general method of ascus formation. 
Dangeard’s (31) separation of the Ascomycetes into two groups 
the so-called ‘‘rectascées’’ and ‘‘curveascées,”’ is not supported 
by the observations of Fraser and Chambers (46), Dale (29, 30); 
Brooks (17), and Blackman and Welsford (12), who report the 
