4 Transactions of the South African Philosophical Society. 



uppermost antipodal nucleus has migrated towards the centre of the 

 embryo- sac, and in Fig. IX. the other polar nucleus has joined it 

 from the micropylar end. It would seem that it is always the 

 antipodal polar nucleus which migrates first, and is then followed 

 by the micropylar polar nucleus. 



Many cases like Fig. IX. have been found, in which the polar 

 nuclei are lying against the wall of the embryo-sac, and a somewhat 

 smaller number of cases in which, as in Fig. X., the polar nuclei 

 (which have now fused to form the definitive nucleus) lie in close 

 contact with the egg apparatus. Fig. X. shows only the egg 

 apparatus and the definitive nucleus. The antipodals can no longer 

 be clearly recognised at this stage. 



The early fusion of the polar nuclei indicates that if a so-called 

 triple fusion occurs in Cassia tomentosa, it is not a simultaneous 

 fusion of three nuclei, but a fusion of the second male nucleus with 

 the definitive nucleus, formed by the previous fusion of the polar 

 nuclei. It has not been possible to investigate this point at 

 present."^ 



Figs. Xl.a, h, c, d, e, f, show various stages of karyokinetic division 

 in the young nucellus. In d, e, and / the chromosomes are seen 

 in the equatorial plane. The number of chromosomes is twelve. 

 Unfortunately the only stage found showing the heterotypic division 

 was not in a condition to show the chromosomes, hence the re- 

 duced number could not be ascertained. There is no reason, 

 however, to suppose that the species is abnormal in this respect, 

 and we may safely assume that the reduced number is half the 

 ordinary number, i.e., six. In conclusion, I have to thank Dr. 

 H. H. W. Pearson for advice and criticism on various points in this 

 investigation. 



Summary. — The chief points noted in the above account are : — 



(1) The innermost megaspore is not functional, as is usually the 

 case, but the innermost but one. 



(2) The antipodal cells, instead of forming a group, lie in a row in 

 a tubular prolongation of the embryo-sac, and, together with certain 

 surrounding cells of the nucellus, appear to function as a conducting 

 tissue. 



(3) The number of chromosomes in vegetative karyokinetic 

 division is twelve. 



Botanical Laboratory, 



South African College. 



* Plants have been kept under observation for some months after fiov^rering 

 commenced, but seed has only been set during January of this year. I hope to 

 investigate the later stages v^hen suitable material has been obtained. 



