Dec. 1S9S.] BOTANICAL SOCIETY OF EDINBURGH 143 



endosperm of Corydalis cava, and Soltwedel makes similar 

 statements with reference to other plants. In reference to 

 these unions, it may be noted that they occur in the 

 degenerate prothallium, and that, according to Strasburger, 

 more than two nuclei may unite to form one. In fact, 

 the phenomenon recalls the repeated nuclear union in the 

 ascus of Peziza Stevensoniana. 



In conclusion, then, according to the hypothesis here put 

 forward, the fusion of the polar nuclei on the embryo-sac 

 is no isolated and unique phenomenon. On the contrary, 

 so far as we know, the process is repeated in every class of 

 the vegetable kingdom where a small degenerate fructifi- 

 cative generation is universally parasitic upon the preced- 

 ing generation. If this interpolated nuclear fusion is 

 evidence of vegetative atrophy, we may be able to employ 

 it as a means of distinguishing between antithetic and 

 homologous alternation of generations, or in ascertaining 

 if there is really such a phenomenon as antithetic alterna- 

 tion of generations among plants. For example, we might 

 find nuclear fusions following the sexual act at an early 

 stage in the development of the sporophyte of Coleochiete 

 or those Bryophyta that possess the simplest sporophytes. 

 On the other hand, should renewed investigation confirm 

 Oltmann's results as to the absence of the double nuclear 

 fusions in the Floridere, we shall feel less inclined to 

 assume that the sporophyte in the family is degenerate or 

 homologous. On the other hand, we may find among the 

 XJstilagineie and Hemiasci types in which deuterogamy 

 plays no part, because vegetative degeneration is not yet 

 sufficiently complete. 



In conclusion, it may be pointed out that the hypothesis 

 here put forward is threefold in nature. According to it 

 — (1) the nuclear fusions discussed are interpolations ; 

 (2) they are concerned with the alternation of generations, 

 and take place in a small fructificative generation ; (3) 

 the small fructificative generation is degenerate, and, in 

 some cases, originally homologous with the other genera- 

 tion. 



If this third proposition were not true, that is, were the 

 fructificative generation neither degenerate nor homologous 

 with the other generation, yet the other two propositions 



