368 W. J. C. Lawrence Cytologia 2, 



in S. vulgaris (Fig. 25 a and b). Afzelius does not give both diakinesis 

 and metaphase for any of the species examined, but no multivalents 

 were found in S. Heritieri (n=30), S, adonifoUus (n=20) or S. cinerea 

 (n=20) which are 12-ploid and octoploid respectively. 



Xanthium (Symons 1926). 



Note groups of two and three bivalents and cf. Dahlia Merckii. 



Hypericum perforatum (Winge 1925 and Nielson 1924). 



Note sizes of associated bivalents. Winge's figure 28 b shows 

 two pairs of associated bivalents in side view. 



Raphanus-Brassica hybrids (n=18) (Karpechenko 1927). 



In these hybrids no bivalents occur at diakinesis. Note the secon- 

 dary association of univalents, in Figs. 13 etc. 



Viola (Clausen 1927, 1928, 1929). 



A great number of the numerous drawings show pronounced as- 

 sociation of bivalents in every degree. It is highly probable that much 

 of this is multivalent association. Diakinesis of V. pinnata (n.= ca 24) 

 however shows a preponderance of bivalents, and the loose association 

 of bivalents in the drawings generally, strongly suggests that secon- 

 dary association is common. See in particular V. striata (n=10) Fig. 5 

 (1929). In this side view of metaphase I there are apparently four of 

 one chromosome type, and three of two others suggesting duplication 

 of types. Note also the sizes of associated bivalents in Figs. 43, 100 

 and 107 (1928). 



Aesculus (Skovsted 1929) 



A. pavia (n=20) has larger chromosomes than A. hippocastanum 

 (n=20). In both these forms described as " diploid " a loose, but 

 obvious association of the bivalents is discernable. A. carnea is a 

 hybrid of these species in which doubling of the chromosome comple- 

 ment has occurred. In Fig. 8 the large chromosomes are associated 

 with large and the small chromosomes with small. Skovsted' s figures 

 provide a convincing demonstration of secondary association, as the 

 associations are too loose to be multivalents. 



Betula (WooDWORTH 1929). 



WOODWORTH supposes the basic number of this genus to be 14, but 

 his drawing of B. japonica (PI. XI Fig. 13) and the cytological situation 



