INTRODUCTION TO THE METHOD 



kind. If, on the other hand, the sets are so different from each other that their chromo- 

 somes cannot pair together, mukivalents are not formed and we have allopolyploidy 

 as in Spartina Townsendii, or to quote a more recently analysed case. Nasturtium uni- 

 seriatum (cf. p. lo). In both types of polyploidy precise phylogenetic evidence can be 

 obtained if the polyploid forms are crossed with others of lower or higher valency and 



Fig. I. Diploid form of Biscutella from central France {B. arvernensis Jord.) showing young fruits, from 

 a herbarium specimen collected in July, near Clermont Ferrand in the Auvergne. Natural size. 



the chromosome pairing in such crosses analysed. This method can indeed be applied 

 to all species or strains closely enough related to be crossed at all, and detailed informa- 

 tion can then be compiled about the relative homologies of their respective chromosomes 

 which may give a surprising insight into past history. Occasionally some similar evi- 

 dence can be obtained more directly by inducing apogamy or parthenogenesis in a 

 normally sexual organism and observing the presence or absence of chromosome pairing 

 in the supposedly haploid set. 



6 



