Dec. 1898.] BOTANICAL SOCIETY OF EDINBURGH 141 



cases under consideration the phenomenon is adapted to 

 enable the parasitic generation to work harmoniously with 

 its parent. The hypothesis is negatived by the fact that 

 in the Ustilaginete, and many of the Uredineae, the spores 

 become detached from the parent before germinating. 



Neither can we suppose that the object of the union is 

 to supply the reproductive cells with nuclei. It is true 

 that where the union takes place in the sporophyte, at any- 

 rate in the fungi, the composite nucleus does supply the 

 spores with their nuclei. But where the union occurs in 

 the gametophyte {e.g. Angiosperms), the composite nucleus 

 is not responsible for the nuclear supply of the egg-cell ; 

 on the contrary, its role is to give nuclei to the vegetative 

 cells. 



Acrain, it might be contended that the nuclear unions 

 on the fungi are essentially connected with a true sexual 

 act, and merely recall the concentration of nuclear material 

 into one spot. This concentration of nuclear material is 

 denoted by Leger and Dangeard's observations on the 

 fertilisation of Spor(fdinia, and by various observations on 

 the oospheres and oospores of Saprolegniaceie and Perono- 

 sporaceffi. In the SaprolegniacecT, according to Strasburger 

 and Hartog, and in Peronospora, according to Wager, a 

 number of nuclei originally present in the oosphere unite 

 to form the single nucleus of the oosphere. But Trow 

 states that there are no such fusions in the Saprolegniaceae, 

 and Mr. Wager has informed me that he is in doubt as 

 to whether the nucleus of the mature oosphere of Perono- 

 spora is, after all, composite in origin. This strained 

 analogy between the behaviour of the nuclei in cells and 

 coenocytes is further weakened by two considerations — first, 

 in the coenocytes it is not a question of the union of two 

 definitely prepared nuclei ; secondly, though there may be 

 repeated nuclear fusions in the Ascomycetes (in Sphacro- 

 theca), several cell-divisions may intervene between the 

 successive acts. 



The next objection which may be urged against the 

 views here put forward is to the assumption that each 

 ascus represents a complete individual. In addition to 

 what has been said earlier in this paper, it may be pointed 

 out that were we to adopt the only other explanation 



