GENETIC INFERENCES 171 



paired at pachytene ; a small reduction leading to rare failure of 

 chiasma formation and metaphase pairing ; a larger reduction, to 

 more frequent but variable failure of pairing ; and a very large 

 reduction to almost invariable failure of pairing. The conditions 

 that determine this reduction are clear from the genetical and 

 cytological evidence of the conditions of pairing just considered. 

 It is that corresponding chromosomes do not agree in structure. 



If mere differences in the arrangement of the same materials are 

 responsible for failure of pairing (and these differences do not in 

 themselves imply differences in the genetic properties of the 

 chromosome set) then frequency of pairing at pachytene is an 

 uncertain indication of the genetic relationship of the parents. 

 And since metaphase pairing is related to pachytene pairing through 

 the formation of chiasmata, in the frequency of which species and 

 varieties differ genetically, the indication becomes even more 

 * uncertain. This conclusion is verified in a general way by a con- 

 sideration of chromosome pairing in hybrid Lepidoptera {cf. 

 Cretschmar, 1927) . Precise examples now show the effect genotypic 

 control can have on the chromosome pairing of hybrids. Thus 

 crosses of Viola nana and V. arvensis with a third species V. tricolor 

 showed almost complete pairing, while when the first two species 

 were crossed less than a quarter of the chromosomes paired (J. 

 Clausen, 1931 c). In crosses of a particular decaploid species of 

 Chrysanthemum with different diploid species different degrees of 

 pairing occur amongst the chromosomes of the decaploid (Table 29, 

 A and B) . The same chromosomes pair or fail to pair under different 

 genotypic conditions. Later we shall find cases where pairing may 

 fail altogether in a homozygous organism (Ch. X). 



But even where there is reason to suppose that genotypic con- 

 ditions are unchanged we cannot infer the degree of differentiation 

 of two species from the pairing in their hybrid without knowing 

 the frequency of chiasmata in all three. This is most clearly 

 shown by the Anas hybrid (cf. Ch. VII and Ch. X). Here the 

 large chromosomes, with a high chiasm a-frequency in the parents, 

 pair regularly in the hybrid. The short chromosomes, with a 

 low chiasma-frequency in the parents, fail to pair in the hybrid. 

 Were all the chromosomes long, no failure in pairing would be 



