622 COMPARATIVE MORPHOLOGY OF FUNGI 



the germ sporangia which have become gonotoconts, however, at first 

 retained their endogenous spore formation but later, because of the long 

 life of the non-privileged sexual nuclei, passed from cleavage to free cell 

 formation and became asci. Thus the ascus would be a Zygomycetous 

 sporangium which, in its character as gonotocont, has become constant 

 in form and spore number. 



According to this conception, then, the step from the Zygomycetes to 

 the Hemiasci would consist in the differentiation of the germ sporangium 

 to an ascus conditioned by the privileging of sexual nuclei, while the 

 structure of the sexual organs themselves and the place of caryogamy 

 remains unaltered. In the step from Hemiasci to Euascomycetes (p. 166) 

 the second point raised for the Phycomycetes, the retardation of caryo- 

 gamy becomes morphologically operative. For some reason caryogamy 

 is more retarded than in Endogone (an intermediate form is not known, 

 with the exception of a suggestion of one in Endomyces Lindneri (p. 142) 

 so that the zygotes not only, as in Endogone, develop to unicellular 

 dicaryotic appendages, but by repeated division of the nuclear pair also 

 develop to much-branched dicaryotic hyphae, the ascogenous hyphae. 

 As fertilization remains fixed in the asci, which therefore serve simultane- 

 ously as zeugites and gonotoconts, plasmogamy and caryogamy are 

 separated in time and space by this insertion of ascogenous hyphae; 

 reproduction is separated in time and space from the sexual act causing it. 



Thus in the life cycle of the Euascomycetes we find a new section, the 

 dicaryophase, whose soma, at first physiologically independent, is embed- 

 ded in the haplont and nourished by it, and in the most extreme case 

 (in Penicillium crustaceum, p. 184) loses its organic connection with the 

 haplont and is parasitic on it. These dicaryotic hyphae in their various 

 types (p. 129) present numerous morphological problems; they may be 

 interpreted as an attempt at a diploid stage which was not possible in 

 this lowest developmental stage of organic life, but which found a 

 physiological, though not morphological, solution in the formation 

 of dicaryons. 



By these multiple, ramose, ascogenous hyphae a whole series of onto- 

 genetic possibilities is suddenly opened to the Euascomycetes; the degree 

 of activity of their gametangia increases in geometric progression, so 

 that, even when the gametangia are uninucleate, several fertilizations 

 are accomplished by one plasmogamy {e.g., Polystigma rubrum, p. 229, 

 in contrast to Eremascus fertilis, p. 139). Similarly, all dicaryons in the 

 polyenergid gametangium retain full activity. The multiple dicaryons 

 of the Pyronema type (p. 332), in contrast to the multiple dicaryons of 

 Albugo Bliti and A. Portulacae, are followed by a corresponding increase 

 in the number of gonotoconts, hence it seems to be significant that in 

 these higher Ascomycetes, the principle of the privileging of a few sexual 

 nuclei may no longer be demonstrated. 



