REVIEW OF FUNGOUS CLASSIFICATION 621 



According to the theory here expressed, the Ascomycetes will connect 

 to the Phycomycetes and particularly to hypothetical Zygomycetes of 

 the Endogone type, in which there arises directly, as the product of the 

 sexual act, a sporangium (the ascus) instead of a zygospore germinating 

 with a sporangium. While in the Phycomycetes, the sexual act leads 

 first to the formation of a resting spore, in the lowest Ascomycetes, the 

 Hemiasci (p. 137), it leads directly to propagation. These lowest Ascomy- 

 cetes still possess in part, like Endogone, polyenergid gametangia in 

 which each nucleus is privileged as active sexual nucleus {e.g., Dipodascus, 

 p. 137; Endomyces Magnusii, p. 143). While in Endomyces Magnusii, as 

 in Endogone, the supernumerary nuclei migrate downwards, in Dipodascus 

 they remain in the mature gametangia beside the functional sexual 

 nuclei; here they persist and at maturation of the zygotes serve vege- 

 tative purposes. 



This circumstance, that in the young germ sporangia of Dipodascus 

 (i.e., in the young asci) numerous sporogenous gametangial nuclei are 

 present, besides the daughter nuclei of the diploid fusion nucleus which 

 are capable of spore formation, is strikingly reminiscent of the relation- 

 ships just described for the phylogeny of the Oomycetes. In the Sapro- 

 legniaceae, even at the time of egg formation, the supernumerary 

 gametangial nuclei have already degenerated, so that (mutatis mutandis 

 as in the sporangia of the Zygomycetes) the whole content of the female 

 gametangium more or less cleaves into oospheres (p. 67). The case is 

 similar for Phytophthora (p. 82). In the higher species of Peronospora 

 and in some of Albugo (p. 84, et seq.), a portion of the supernumerary 

 nuclei remains at the periphery of the female gametangium, so that the 

 contents of this gametangium separate as a central egg cell, which serves 

 the true sexual function, and into a peripheral periplasm that performs 

 only vegetative tasks. A convergent development has apparently 

 occurred in the Zygomycetous sporangia which have become gonotoconts. 

 If the contents of these sporangia divide directly by cleavage, as occurs 

 in the ordinary sporangia of the Zygomycetes, the remaining, non- 

 privileged, gametangial nuclei would pass into the spores; that can only 

 be avoided if each of the sporogenous nuclei forms a membrane in its 

 environs which leaves unused as (vegetative) periplasm the remaining 

 protoplasm and the gametangial nuclei. In this sense, the ascus of the 

 Ascomycetes and the oogonium of the Oomycetes may be regarded as 

 analogous structures. 



One must, then, apparently imagine that the sporangia of the here 

 discussed (hypothetical) Zygomycetes have split into two types, just as 

 occurred in another direction in the Choanephora-Blakeslea group (pp. 

 100-105). The ordinary vegetative sporangia retained their Zygo- 

 mycetous character (division into spores by cleavage) and developed gradu- 

 ally to exogenous forms (Cunninghamella-Syncephalis- Aspergillus group) ; 



