140 TRANSACTIONS AND PROCEEDINGS OF THE [Sess. LXiii. 



act in the ancestors, in fact, it is superposed upon and 

 subsequent to the sexual act. Bearing in mind the analogy 

 to the sexual act, we may describe these nuclear unions 

 as being cases of deuterogamy. 



In seeking to give a deeper and more final explanation 

 of deuterogamy we are met with the initial difficulty that 

 we are quite in the dark as to the significance of the 

 nuclear union taking place during a sexual act. Further- 

 more, the accentuated difficulty which assails us in our 

 endeavour to conceive of the utility of a fusion between 

 two nuclei formed in one cell (basidiuai, ascus, or spore), 

 also comes upon us in considering the sexual unions 

 between two adjoiuing- cells belonging to one individual, or 

 in considering the coalescence of the two nuclei in the 

 young auxospore of Syncdra that has been produced 

 without conjugation with another individual. 



The suggestion that the object of the nuclear union is 

 to double the cjiromosomes is negatived by the occurrence 

 of a double union in the Asconiycetes, and by the fact 

 that in Angiosperms the chromosomes in the endosperm 

 sometimes preserve the half number. In fact, one can 

 only fall back upon the idea tliat, viewed as mere 

 physiological problems, tlie coalescence of two nuclei, 

 both in the sexual act \i'p. Strasburger (;5)] and in 

 deuterogamy, is bound up witJi the nuclear divisions 

 which precede (or possibly succeed) those fusions. The 

 division of a nucleus may afford the latter an opportunity 

 for changing its constitution, for ejecting intra-nuclear 

 protoplasm, and for admitting extra-nuclear protoplasm. 

 And tliere is much significance in the fact that in all 

 the most carefully investigated cryptogamic examples — 

 Uredineffi, Ascomycetes, and Basidiomycetes — the nuclei 

 before coalescing apparently do eject material which is 

 described as being kinoplasmic or nucleolar in nature. 



For the purpose of clearness, the consideration of 

 difficulties and alternative explanations has been post- 

 poned to this stage. A few difficulties and other hypo- 

 theses are now briefly dealt with. 



It might be tliought that just as nuclear unions have 

 been suggested between the cells of graft and scion in order 

 to permit of the co-operation of tlie two symbionts, so in the 



