I93I The secondary association of chromosomes 3g9 



in Corylus point to a basic number of 7. B. papyrifera and the tetra- 

 ploid var. cordifolia would be therefore decaploid and octoploid respec- 

 tively. WooDWORTH has a significant note on the association of univa- 

 lents in pairs in D. japonica var. mandshurica (n=28). This may be 

 autosyndesis or secondary association. 



Corylus (WoODWORTH 1929). 



The chromosome number of all the species is n=14. " Throughout 

 the forms there occurs a fusion of one, two or three pairs of bivalent 

 chromosomes, thus often causing the haploid number to appear to be 

 lower than fourteen." Figs. 1, 21 and 27 (diakinesis) show 1, 2 and 3? 

 quadrivalents respectively. Secondary association of the bivalents is 

 seen at metaphase. Woodworth concludes ' Corylus exhibits no poly- 

 ploidy "! Note the grouping of bivalents in Alnus (Figs. 48-50). 



Solanum species (Longley and Clark). 



The drawings of metaphase in diploid species (n=12) suggest that 

 6 was the original basic number of the Solanaceae (cf. Fig. 1 d). Note 

 the association of univalents in Fig. 6 a and b. 



Nicotiana. 



Brieger, (1928). Note sizes of associated chromosomes. 



Chipman and Goodspeed (1927). Note secondary association in the 

 haploid N. purpurea (PI. 5, Fig. 15). The authors remark that "the 

 occurrence of ' bivalents ' at I M is not a reflection of pairing before 3rd 

 contraction, because no evidence can be found of bivalents at diakinesis 

 or earlier" (the italics are mine). Their suggestion that the close 

 association of the chromosomes during 3rd contraction produces varying 

 amounts of adhesion between the univalents is interesting, but, as they 

 remark, if it were merely such a mechanical effect then larger groups 

 of more than two chromosomes would be expected. The largest groups 

 found however were of 3 univalents only. It seems probable that the 

 N. purpurea haploid, like the Brassica-Raphanus hybrids, provides 

 direct evidence of the secondary association of homologous univalent 

 chromosomes. 



Comparison should be made between the drawings of Brieger, 

 Goodspeed and Clausen, Goodspeed, Clausen and Chipman, and 

 Chipman and Goodspped to see the apparent association of univalents 

 in all cases. Cf. also Karpechenko (1927, 1928) and Woodworth 

 (1929) on Betula japonica var. mandschurica. 



