The Behavior of the Chromosomes as Studied through Linkage. 255 



he anything in the chromosome view, these must all be located in the 

 same or homologous chromosomes. Morgan ('10c, '11a) did in fact 

 show that there must be an interchange of factors between the two sex 

 chromosomes in the female. A female which had received a sex chromo- 

 some bearing genes for red eyes and for rudimentary wings from her 

 father, and one bearing genes for white eyes and for long wings from 

 her mother, produced gametes of all four sorts: red long, red rudiment- 

 ary, white long, white rudimentary. In this way the Correns-de Veies 

 idea of interchange, or as we now call it, crossing over, was demon- 

 strated to be correct in at least one case. 



Later (Morgax 'lib) it was discovered that this interchange is 

 not always free. Thus, a female bearing genes for red eyes and for 

 gray color in one sex chromosome and for white eyes and yellow color 

 in the other produces gametes of the kind resulting from crossing over 

 (i.e., red yellow and white gray) only very rarely — about one of them 

 to a hundred gametes bearing chromosomes in which these genes have 

 not interchanged. Moreover, the proi)ortion of cross-overs is fairly 

 constant for any given pairs of allelomorphs, but may be very different 

 for other pairs. Many illustrations of this are given in this paper. 

 These experiments demonstrated the correctness of Correxs' and Lock's 

 suggestion that linkage may be due to a failure of interchange between 

 homologous chromosomes. Lock had said that there was no cytological 

 evidence as to how this interchange or failure of it may be brought 

 about; but that lack has since been remedied by Janssexs ('09). This 

 observer, and others later, finds that during the period just before the 

 maturation divisions homologous chromosomes twist around each other 

 like the strands of a rope: and ajiiiarently do ndt always simply untwist, 

 but may fuse together at certain points, and then pull apart in sucii a 

 way as to bring about an interchange of homologous regions (see Fig. 2). 

 Morgan ('lib, 'lie) has applied this conception to his experimental results, 

 using Eoux's idea of linear arrangement, and supposing the strength 

 of linkage to be dependent upon the distance apart of the genes involved. 

 Th(i writer (Sturtevant '13 and the pi-esent paper) has attempted to 

 analyze further this idea that strt'iigth of linkage is merely a mechanical 

 result of relative position in the cln\)mosomes. To me it seems that the 

 study has given strong confirmation for Morgan's hypothesis. That 

 the relations existing between the factors of a given group may be ex- 

 pressed by arranging them in a linear series, and that this linear series 

 offers a means of accurate prediction for untried crosses — in short, that 



