1963] 



TAXONOMY OF LUZULA SUBGENUS PTERODES 



281 



spike could also develop by a further reduction until only a single peduncle 

 remains. 



Some information concerning the interspecific relationships of the species 

 can be determined by hybridization and the fertility of the F 1 progeny. Breed- 

 ing experiments conducted by Nordenskiold (1957) show to some extent the 

 relationship of the species in the subgenus Pterodes by the amount of fertility 

 of the Fi hybrids. However, even with this information it is usually impossible 

 to tell in which direction evolution is taking place, because most of the crosses 

 produced sterile Fi hybrids. 



The size of the chromosomes and polyploidy are probably the most important 

 characters for determining phylogenetic relationships within the genus. This is 

 due to the diffuse nature of the centromere which is characteristic of the chromo- 

 somes of this genus. In contrast, in most plants and animals the chromosomes 

 have a localized centromere which is necessary for the movement of the chromo- 

 somes to the poles of the spindle at anaphase of mitosis and meiosis. X-ray ex- 

 periments have shown that fragments of chromosomes which lack a centromere 

 region are usually eliminated in the course of a division cycle. Therefore, the 

 localized centromere was commonly considered to be a universally occurring 

 structure of the chromosome until Schrader (1931) demonstrated the existence 

 of the diffuse or spreading nature of the centromere for the X-chromosome of 

 the insect Protortonia primitiva. The existence of diffuse centromere activity in 

 plants was not known until Malheiros and de Castro (1947) discovered its oc- 

 currence in the family Juncaceae. They observed that during somatic division 

 the chromosomes are arranged at right angles to the axis of the spindle during 

 metaphase and at anaphase the chromosomes separate parallel to each other. 

 11 was also observed that the chromosome ends are slightly turned to the poles, 

 in sharp contrast to the normal V-shaped configuration characteristic of chromo- 

 somes with localized centromeres. During both divisions of meiosis the chromo- 

 somes perform the movement to the poles in the same manner as in somatic 

 divisions. 



Due to the diffuse nature of the centromere in the genus Luzula, chromosome 

 size as well as chromosome number can be used in determining phylogeny. About 

 one-third of the species determined are characterized by a somatic chromosome 

 number of 12. These chromosomes are essentially the same size in the cell and 

 are about 1-2 p long and 0.5 fx wide when on the metaphase plate. The chromo- 

 some number of the other species of the genus are more numerous. The origin 

 of these high numbers has not been determined with certainty, but probably 

 they evolved in two different ways. Nordenskiold (1951) reports that within 

 the genus three different chromosomal deviations from the standard pattern 

 (2n = 12) can be traced. These deviations usually form a polyploid series and 

 all numbers reported conform to one of these series. 



The first type is an ordinary polyploid series in which the standard chromo- 

 some complement (2ft = 12) is doubled twice, three or four times to form tetra- 

 ploids, hexaploids or octoploids. These polyploidal species having chromosomes 

 of about the same size as those of the standard chromosome have been referred 

 to as the AL-type by Nordenskiold (1951). According to her, it seems very 

 unlikely that all of the species arose by a repeated doubling of the chromosomes 

 in the same manner as an ordinary polyploid series. Even though the chromo- 

 some numbers of these species conform to the multiple series of the genus, the 

 difference in size is too large to make this possible. Therefore, Nordenskiold 

 (1951) suggested the possibility of another series. 



