Varietäten, Descendenz, Hybriden. 185 



ölten recorded Variation in the chromosome number, and possibly 

 also of the mendelian phenomenon oi' coupling. 



Agnes Arber (Cambridge, England). 



Gates, R. R., Pollen Formation in Oenotheva gigas. (Ann. Bot. 

 XXV. p. 909—940. pl. 67—70. 1911.) 



At the beginning of synapsis there is a rather sudden increase 

 in the volume of the nucleus, there being in some cases a disten- 

 sion of the nuclear membrane and in other cases a rupture of the 

 membrane and aecumulation of karyolymph in the cytoplasm. In 

 some fiowers, but probably not in all, there is an extrusion of 

 chromatin from the nucleus of one mother-cell through cytoplasmic 

 connexions, into the cytoplasm of an adjacent mother-cell. After 

 this, the nucleus moves back to the centre of the cell, and various 

 other transformations oecur. As the chromosomes pass to the poles, 

 from the equatorial plate of the heterotypic spindle, they frequently 

 undergo a split, which in some cases gives evidence of being trans- 

 verse, in others longitudinal. 



It is considered most probable that the mutant O. gigas, vvith 

 the tetraploid number of chromosomes, originated through a suspen- 

 ded mitosis, either in the fertilized eggs or in the megaspore 

 mother-cell. Agnes Arber (Cambridge, England). 



Harris, J. A„ On the Selective Elimination oecuring du- 

 ring the Development ofthe Fruits of Staphylea. (Biome- 

 trika, VII. p. 452—504. 1910.) 



The author has made observations on the selective elimination 

 of the fruits between the time of flowering and maturity in Staphy- 

 lea. The problem is therefore one of the selective elimination of cer- 

 tain types of organs produced by individuals, as distinet from the 

 selective elimination of individuals themselves. 



Ovaries with a relativefy low number of ovules are more exten- 

 sively eliminated than those with higher numbers; the raean of the 

 population after elimination being about 7 or 8 per cent higher than 

 that of the eliminated individual organs. 



As might be expected from the foregoing, the variability both 

 of the eliminated ovaries and of the series remaining after elimina- 

 tion is less than that of the original population. 



Ovaries remaining after elimination are more radially symmetn- 

 cal than those which are eliminated. 



Ovaries with one or more loculi containing an „odd" number of 

 ovules are more likely to be eliminated than those in which all the 

 loculi contain an even number. 



Dimerous ovaries seem to be less likely, tetramerous ovaries to 

 be more likely to develope to maturity than the normal trimerous 

 ones, but further data are needed on this point. 



Radial symmetry and the composition of the fruit with regard 

 to the presence of loculi containing u odd" or "even" numbers of 

 ovules are not independent, but correlated characters. It may there- 

 fore be possible that one of these characteristics has comparatively 

 little significance in determining whether or not an ovary shall con- 

 tinue development, but reasons are adduced for believing that both 

 characteristics are to some extent of independent significance in this 

 respect. 



