424 BREAKDOWN OF GENETIC SYSTEMS 



" translocation " (Navashin, 1931 ; Lewitsky et alii, 1931, 1934) • 

 Apart from these changed but characteristic chromosomes there 

 occur others of a kind that are not normally found. These are 

 commonly of four kinds : ring chromosomes, branched chromosomes 

 (with lateral trabants) and the two types that are characteristically 

 produced by crossing-over in inversions at meiosis, viz., dicentric 

 and acentric chromosomes. As we should expect, the acentric 

 chromosomes are unaffected by the metaphase congression and 

 anaphase separation, while the dicentric chromosomes divide, 

 either regularly to give daughter chromosomes like themselves or 

 criss-cross to give double bridges which break ; the difference 

 depends on whether the chromatids are parallel or have half a coil 

 between the two centromeres (Mather and Stone, 1933). A third 

 type of division is found where the centromeres lie farther apart, 

 viz., a complete coil between them. The two centromeres of each 

 chromatid then pass to the same pole, but the chromatids, each 

 M-shaped at anaphase, are interlocked in the middle loop of the 

 M's (Roller, 1934, Plate XII). 



The acentric chromosomes may be supposed to result from 

 simple breakage, but some of them will arise from as3nnmetrical 

 interchange, which must be responsible for the dicentric chromo- 

 somes and the ring chromosomes. 



From the later results of these structural changes of whatever 

 kinds, Navashin (1932) concluded that there was never any change 

 in the number of centromeres and that centromeres were therefore 

 self-perpetuating bodies. This conclusion is verified by the 

 observations of the effects on the immediately following divisions 

 by Mather and Stone ; the two centromeres of one chromosome 

 are independent of one another, while the chromosome without a 

 centromere is deprived of the means of autonomous movement. 

 It is usually lost in the cytoplasm at telophase. 



We have seen from the comparison of mitosis and meiosis that 

 the chromosomes must be supposed to divide after telophase 

 always, and in mitosis, but not in meiosis, before the prophase. 

 It seems probable, therefore, that their division occurs at the end 

 of the resting stage in mitosis, possibly even determining the onset 

 of prophase in certain circumstances (D., 1932). We cannot find 



