328 INHERITED LINKAGE VARIATIONS 



SIGNIFICANCE OF MAP DISTANCE. 



It has often been pointed out (e. g., Sturtevant, 1913, p. 49; Morgan, 

 Sturtevant, Muller, and Bridges, 1915, pp. 67-68) that 1 per cent of 

 crossing-over must not be supposed to represent the same actual 

 morphological distance in different chromosomes or in different regions 

 of the same chromosome. Actual distance is evidently an important 

 factor in the result. Other things being equal, chromosome sections 

 of equal length will give equal percentages of crossing-over; but in no 

 case can we be certain that "other things" are equal. The terms 

 ''distance" and "percentage of crossing-over" have unfortunately 

 been sometimes used almost as though synonymous, and confusion 

 has perhaps resulted. But it has been recognized from the beginning 

 that different regions might show different frequencies of crossing-over 

 for the same actual length of chromosome. 



The results presented in this paper show conclusively that this is 

 the case, as has already been stated (Morgan, Sturtevant, Muller, 

 and Bridges, 1915; Muller, 1916; Sturtevant, 1917). They show that 

 even in the same chromosome pair the percentage of crossing-over 

 shown by different regions is not only not always the same, but is 

 not necessarily even proportional. For example, while S' b remains 

 approximately 40.0, b c may be either 23.0 (neither Cm nor C IIr 

 present), or 7.5 (heterozygous C IIr ). 



LINEAR ARRANGEMENT OF GENES. 



The strongest evidence for the linear arrangement of genes is that 

 derived from crosses in which more than two loci in the same chromo- 

 some can be followed. The method of seriating the loci on the basis 

 of such information has been described in detail elsewhere (Morgan, 

 Sturtevant, Muller, and Bridges, 1915; Sturtevant, 1915; Morgan and 

 Bridges, 1916), so need not be discussed here. When the linkage 

 values are changed the question arises: Is the sequence of genes 

 affected? It has already been shown (Bridges, 1915; Plough, 1917) 

 that this sequence is not altered when the amount of crossing-over 

 is changed by age or by temperature. In the case of the genetic 

 changes reported here, the evidence presented in tables, 1, 3, 10, 

 12, 14, 17, and 19 shows that the sequence found in "normal" females 

 is maintained. There are just three cases in which the data, uncor- 

 rected by other data, might lead us to assign a different sequence. 

 These three cases may now be taken up in turn. 



(1) In the case of 7 ~ only one cross-over between 6 and 



6 p r v g a r s p 



p r was obtained; and that was also a cross-over between p r and v g . 

 This would lead us to suppose the sequence to be p r b v g , were no 



