INDUCED CHROMOSOMAL ABERRATIONS IN ANIMALS 1183 



of chromosomes while the latter should at least in certain cases do so. 

 The question arises whether the genetic methods used for the study of 

 chromosomal rearrangements are dependable in the sense that the con- 

 clusions arrived at by these methods can be justified by cytological 

 observations. 



Stern (126, 127, of. Muller, 74) was first to discover cytologically 

 alterations of the chromosome structure produced by translocations 

 between the X- and the F-chromosomes of Drosophila. He concluded 

 on the basis of genetic data that in a certain stock a fragment of the 

 F-chromosome was attached to the spindle-fiber end of the Z-chromo- 

 some, near the locus of the gene bobbed. An investigation of the 

 chromosomes has shown this conclusion to be valid. Muller and Painter 

 (80), Painter and Muller (98), Dobzhansky (24, 27), and many other 

 authors on Drosophila, and Nabours and Robertson (84) on Apotettix gave 

 further examples of the same kind. In a series of translocations, genetic 

 data suggested that fragments of certain chromosomes were attached to 

 other known chromosomes. As expected, it was found cytologically 

 that some chromosomes were shorter than normal, and that other 

 chromosomes were increased in length by about the same amount of 

 material which was subtracted from the former. 



These results are interesting not only because they give an additional 

 proof of the soundness of the fundamental assumptions of modern 

 genetics, but especially because they furnish a method whereby certain 

 problems can first be attacked experimentally. Among such problems 

 is that of the distribution of cross-overs in the chromosomes, and that 

 of the nature and significance of the genetic maps. 



Genetic maps, first obtained for Drosophila melanog aster by the 

 Morgan school, are known at present for the chromosomes of a series 

 of species of animals and plants. Genetic maps show the linear order 

 in which the genes are arranged within the chromosomes, and also the 

 distances from one gene to the other, expressed in terms of the frequencies 

 of crossing over between them. The information used in construction 

 of genetic maps consists of the data on the degree of linkage exhibited 

 by genes in various combinations. Consequently, genetic maps are 

 merely a most condensed summary of the available statistical data on 

 the behavior of genes in different crosses. 



The relative distances between the genes on the genetic maps may 

 correspond to the actual spatial relationships in the chromosomes only 

 provided the frequency of crossing over per unit distance is constant in 

 all parts of the chromosomes. This proviso has never been proved; 

 hence it has always been realized that the genetic maps may give a 

 distorted conception of the relative lengths of the different parts of the 

 chromosomes (Bridges and Morgan, 19; Morgan, Bridges, and Sturtevant, 

 71; Muller, 72). Obviously, if crossing over per unit length takes place 



