1904] DAVIS—SEXUAL ORGANS IN PLANTS 261 
precise statement of the evolutionary tendencies of coenogametes, 
but certain factors may be considered, of which the principal ones 
seem to be cooperative in both the Phycomycetes and Ascomycetes. 
Assuming that coenogametes may have arisen at various times inde- 
pendently of one another, and from an ancestry at approximately 
the level of isogamy or slightly above it, their evolution might proceed 
along three or more divergent lines. They all agree in having very 
numerous potential gamete nuclei, and there is strong evidence from 
the processes of gametogenesis in Saprolegnia, the Peronosporales, 
Pyronema, and Vaucheria that these are under conditions which 
demand extensive nuclear degeneration. Consequently the evolu- 
tionary tendencies are largely concerned with the disposition of 
superfluous nuclei and seem to present the following possibilities. 
I. General nuclear degeneration may result in the survival of a few 
gamete nuclei in relation to coenocentra and the development of a 
imited number of eggs, as in the Saprolegniales. 
II. Superfluous nuclei with some cytoplasm may be differentiated 
‘$4 periplasm, with functions to perform in laying down portions of 
the spore wall, which conditions accompanied by numerical reduction 
of the nuclei in the ooplasm give the general tendencies in the Peron- 
osporales. There js apparently presented in Araiospora (KING ’o3) 
* modification of the habits in the Peronosporales, since the periplasm 
in this form develops a cellular envelope around the spore. 
— the Ascomycetes we have a much wider range of conditions, 
= portionally much less knowledge of the forms, so that the 
pag : os of evolutionary lines becomes very speculative. How- 
hin ag Hous Protoplasm with nuclei is used here to form acces- 
ae i . such as the conjugation tube of Pyronema and the 
mbection, wh: eer There is probably also numerical nuclear 
Sehiivite. oe culminate in uninucleate gametes, as in 
line aus a : € multicellular trichogynes. and archicarps of the 
logical aboulbeniaceae present some very difficult morpho- 
Rh reams with possible- relations, however, to conditions in 
establish ag ceae, especially should further study in the latter group 
Presence of multinucleate sexual organs. 
S of the paragraph above must of course stand the 
= Suggestion 
est extensi * . . 
Nsive Mvestigations on many more forms and with refer- 
