246 CHROMOSOMES IN HEREDITY: MECHANICAL 



could at once be described as genes, but this is not so. A gene is 

 the unit of crossing-over, and therefore the atom of inheritance. 

 Until we consider more in detail therefore how crossing-over occurs, 

 we cannot tell how big the gene may be ; we cannot even tell 

 whether the same particle always behaves as a unit. 



We can, however, proceed to treat the chromosome as a string of 

 particles, provisionally described as genes, whose changes are 

 responsible for the differences in hereditary characters that we know 

 as mutations. 



5. Mutation. Experimental breeding has shown that the genes 

 must be capable of at least two kinds of change : those affecting 

 their individual properties (gene or point mutation) and those 

 affecting their numerical proportions and arrangement (deficiency, 

 translocation, etc.). Corresponding structural changes in the 

 chromosomes have been shown (Chs. Ill, V). The existence of 

 qualitative differences between genes in the same complement is 

 evidence of gene-mutation on an evolutionary hypothesis, for all 

 the different genes found to-day cannot have arisen independently 

 at the beginning of life. 



Some of the principal evidence for regarding the chromosomes as 

 containing the hereditary materials has therefore already been 

 given. The most unequivocal test of the chromosome theory, as 

 indeed of any theory, is by the verification of prediction. It is 

 possible to base prediction of chromosome form and behaviour on 

 observations by genetical methods, and vice versa, and these 

 predictions can afterwards be tested. Thus it is possible in the 

 case of certain flies mosaic for male and female characters to say 

 that one X chromosome has been lost in the course of a somatic 

 mitosis (Stern, 1927). It is possible to say in the case of seedlings 

 such as those raised from the cross Rubus rusticanus by R. thyrsiger 

 or from natural seed of the Raphanus-Brassica hybrid (Ch. VII) 

 that some have one set of chromosomes more than others. These are 

 predictions based on genetical evidence and they have been verified 

 cytologically. On the other hand, it is known that meiosis occurs 

 in the formation of spores of mosses and ferns, and that the gameto- 

 phytes raised from them are haploid : they should therefore show the 

 I : I segregation in haploid characters which Mendel supposed 



