328 CHROMOSOMES IN HEREDITY : PHYSIOLOGICAL 



as described in the previous chapter in relation to secondary 

 structural change (c/. also p. i8i). 



It may be thought remarkable that pairing (and crossing-over) 

 should occur between parts of chromosomes in the haploid, although 

 these parts never seem to pair in the diploid, especially since in 

 CEnothera this pairing is often terminal, proving that the ends of 

 some of the different members of the haploid set are homologous. 

 How does this agree with the notion that the parts of the chromo- 

 somes are qualitatively differentiated ? A comparison with the 

 observations of differential affinity in polyploids shows the explana- 

 tion (Ch. VI). It is known that chromosomes which rarely or never 

 pair, in an allotetraploid, may pair regularly in its diploid hybrid 

 parent or its parthenogenetic diploid progeny. Thus it is not 

 surprising that chromosomes which never pair in the diploid should 

 occasionally pair in its parthenogenetic haploid progeny for the same 

 reason. Two " non-homologous " chromosomes of the haploid set 

 may consist of a series of particles (or genes) ahcdefghkl . . . and 

 amnopqrs ... In the presence of identical mates, i.e., with the same 

 linear sequence of particles, the a particles of different chromosomes 

 will never pair. In the haploid they will very often pair at pachytene 

 and occasionally form a chiasma which will preserve their pairing at 

 metaphase. 



These observations are in agreement with the theory of mutation, 

 by crossing over, outlined elsewhere (Chs. VII and IX), which 

 demands that in CEnothera, as in other organisms, small sections 

 should be homologous in different parts of the same haploid com- 

 plement and of different complements. This would mean that 

 there had been a considerable amount of reshuffling of small pieces 

 of chromosome in the history of the species, the primary changes 

 being inversion and interchange, and the secondary changes, 

 reduplication and deficiency. 



Every gene or series of genes must be supposed to have had a 

 beginning at some time when it would have been singly represented 

 in any chromosome complement. Reduplication of genes therefore 

 means that there has been a change of proportion, which if successful 

 becomes secondary balance. All chromosome complements are 

 probably secondary in this sense, for such change is perhaps the 



