238 



THE BEHAVIOUR OF POLYPLOIDS 



ship if they could be measured sufficiently accurately. This is of 

 practical importance in the analysis of polyploid}^ Most ordinary 

 cases are clear enough from purely numerical considerations, but 

 where no species is known with a lower basic number or where the 







/ 







^P:t .?!*• 





AAA A 



DD 





•.V 



Fig. 8o. — Secondary pairing at meiosis in the pollen mother-cells. 

 h and e to A, first metaphase ; /, first anaphase ; a, c and d, 

 second metaphase. a, Euphorbia scordifolia {2x = 20). h, c, 

 E. granula {4X = 40). d, Rubus " Loganberry " {6x = 42). 

 e,f, Dahlia coccinea {4X = 32). g, h,j, D. Merckii {4X -f- 4 = 36). 

 k, I, Pyrus Mains {4X + 6 = 34). a to c, after Hagerup, 1932 ; 

 d, from Crane and D., 1927. e to j, from Lawrence, 1929 and 

 1931. k and /, from D. and Moffett, 1930, and Moffett, 1931. 



number is not a simple multiple, the evidence of secondary pairing 

 becomes of use. The distribution of chromosomes on the equatorial 

 plate may be used to show relationships between them, or, more 

 exactly, between some of their constituent parts, which are beyond 

 analysis by the two other methods — comparison of shape at mitosis 



