334 CHROMOSOMES IN HEREDITY: PHYSIOLOGICAL 



have made use of new agents of analysis : the combination of 

 X-ray breakage with the examination of the results both pheno- 

 typically and by salivary gland study. They have shown that the 

 smallest unit of X-ray breakage corresponds with the smallest 

 chromomeres detectable in ultra-violet light (cf. Ch. X). They 

 have also shown that this unit is smaller than the unit of crossing- 

 over and that changes inside the crossing-over unit are sometimes 

 due to structural changes which have a position effect. What 

 relation there may be between this classification of mutations by 

 their chemical causes and that by their physiological effects (as 

 classified by Muller, 1932) is not yet clear. 



These conclusions, along with some others, show that a one-to- 

 one correspondence of particles and differences can no longer be 

 assumed. In particular they make it impossible to express all 

 variation in terms of units directly derived from its material basis, 

 the genes. Any advance in the theoretical treatment of variation 

 must therefore wait on a more exact definition of the mutual 

 relationships of particles and changes. 



In spite of the complexity of these relationships we can assume 

 without hesitation that an intra-molecular and therefore intra- 

 genic change precedes and conditions all more complicated kinds 

 of change. It follows that intra-genic change must inhibit pairing 

 of the differing genes. Thus pairing of genes may be prevented 

 at pachytene by either of the two changes which are ordinarily 

 described genetically : intra-genic change will prevent pairing of 

 two genes, quantitative change will prevent pairing of a larger or 

 smaller group of genes according to the number shifted (which may 

 amount to the whole chromosome). Therefore quantitative or 

 structural change will be a vastly more potent means of preventing 

 pairing, and the failure of pairing in hybrids may be taken as an 

 indication of the part that structural dissimilarity between corre- 

 sponding chromosomes has played in the differentiation of their 

 parents. Further, it follows that allelomorphism essentially 

 depends on a correspondence of position in the chromosome rather 

 than on the chemical relationship which is usually associated 

 with such a correspondence. 



I 



