110 GENETIC CYTOLOGY IN SWEDEN 



mentioned Drosera^bastard, this irregularity proved to be merely apparent, as, dur« 

 ing metaphasis, 10 regularly formed gemini appeared in the equatorial plate, while 

 10 unpaired or single chromosomes were distributed over the spindlesfigure. Start* 

 ing from Montgomery's well»known hypothesis with regard to the chromosome* 

 pairing in the heterotypic division, the above»mentioned chromosome»arrangement 

 is thus explained: that 10 D. rotundif olia=chromosomes become paired with 10 

 D. longifolia, and the remaining /ong^//b/ia«chromosomes constitute the 10 unpaired 

 ones. The result finally is the production of germ«cells of greatly varying chro* 

 mosome*numbers, as, of course, the unpaired chromosomes are distributed very ir« 

 regularly on the daughter nuclei. 



A large number of other cases with bastards of the type mentioned, x + 2x 

 or 2x + y, where y signifies a smaller number of chromosomes than x, has, later 

 on, been investigated by the writer in other species, and have proved to follow 

 the same Drosera=scheme. 



By Mendel's well«known crossing^experiments with Hzeracium=species, there 

 was discovered the remarkable fact that bastards were constant. The experiments 

 made, later on, by Ortenfeld showed that this depended on the occurrence of 

 apomixis or parthenogenesis in Fi and that the hybrids would segregate, when crossed 

 back with the parents. On the other hand, Ostenfeld obtained, in some of his 

 crossing*experiments, a segregation already in Fi, for instance, in a single cross 

 between Hieracium auricula (sexual species) and H. aurantiacum 29 hybrid indi* 

 viduals, all of which differed from each other. The writer has shown (1917) that 

 the germ^cells in one of the parent plants showed very different number of chro« 

 mosomes, which resulted in a greatly varying chromosome^arrangement in the 

 bastard individuals. 



In the section Archieracium the sexual species have 9 chromosomes in the 

 germ cells; among the apogamous species the number of chromosomes in the 

 somatic nuclei is 27, thus proving, that these species must be considered as hybrids 

 between parent plants with 9, resp. 18 chromosomes in the germ«cells. 



An interesting fact in Hieracum is, that the apogamous species of Pilosellae 

 are not absolutely apogamous but form both sexual embryo«sacs, which have 

 proceeded from a reduction«division, and embryo<sacs with the diploid number of 

 chromosomes, which have proceeded from a vegetative cell in the vicinity of the 

 sexual one, and displacing this latter during the course of its development. 



A very peculiar circumstance was observed in the case of a bastard between 

 H. excellens and H. aurantiacum where the germ*cells of the parent=plants have 

 each 18 gemini«chromosomes and a number of unpaired ones, while the bastard 

 in the heterotypic division shows over 22 gemini plus a number of unpaired 

 chromosomes. This depends thereupon, that some of the unpaired chromosomes 

 in each germ*cell are homologous and therefore form gemini. A similar condition 

 of things has also been shown by Holmgren (1919) in his important investigation 

 of genus Erigeron. 



Of extreme importance is a newly published work by Tackholm (1920) on 

 the cytology of the genus Rosa. The purpose of that study was to learn to what 

 extent the use of cytological methods of research could be employed in the 

 elucidation of the problems that appear in the genus, distinguished by its enormous 



