I93I The secondary association of chromosomes 375 



and genera with high basic numbers will be found to be polyploid or 

 secondarily polyploid (cf . Salicaceae, Betulaceae, Moraceae, Magnolia- 

 ceae, Vitaceae etc). The absence of low numbers in these and other 

 genera and orders would seem to imply that the polyploid forms had a 

 distinct advantage over the ancestral diploid forms, all of which are 

 now extinct. Thus it may be true that the genus Solanum, in which 

 no species is found with less than 24 chromosomes, arose from n=6 

 forms. The occurrence of related genera with less than 24 chromo- 

 somes (viz. Petunia n=7, Nicotiana n=9 and n=10) seems to support 

 such a conclusion, though it cannot be denied that chromosome fusion 

 would also account for these lower numbered forms. The related 

 orders Labiatae and Scrophulariaceae both have a low basic number. 

 The apparent association of the 12 chromosomes of the haploid sets of 

 Solanum species figured by Longley and Clark (1930) possibly indicates 

 such a constitution for Solanum as outlined above. 



If the tempo of differentiation, by point mutation or by structural 

 reorganisation of the chromosome complement, is more or less uniform 

 throughout the plant world then, chromosome size apart, we might 

 expect the oldest allo-polyploid species to show the least secondary as- 

 sociation. Where differentiation is complete and the individual is vir- 

 tually diploid then no association would be expected. 



Fragmentation, reduplication, translocation and segmental inter- 

 change have undoubtedly played an important part in the evolution of 

 new forms, and because these phenomena do occur we cannot preclude 

 the possibility of secondary association between chromosomes which 

 have been structurally reorganised. Segmental interchange has now 

 been demonstrated in six diploids " and in view of the fact that ring 

 formation is regular in these cases it follows that secondary association 

 will occur under similar and favourable conditions. 



The study of meiosis in the " haploid " derivatives of polyploids is 

 becoming increasingly important. This is especially true in regard to 

 the allo-polyploid, for in the absence of the homologues which regularly 

 pair with their kind, the more remote affinities of the constituents of 

 the haploid set are freely expressed. The behaviour of univalents may 

 also be observed in hybrids in which a certain number of chromosomes 

 are without mates or fail to form chiasmata with their homologues. 



In meiosis of hybrids the distribution of the univalents on the 

 spindle is usually entirely fortuitous, except that they are sometimes 



1) Pimis, Zea, Campanula, Briza, Oenothera and Datura. 



