144 PHYSICAL BASIS OF HEREDITY 



curved (c), and speck {s2:)) are indicated in their relative 

 locations, i.e., spaced in proportion to the usual amount 

 of crossing over between them. Correspondingly, the 

 short second line is based on the crossover relations of 

 these factors when the female is heterozygous for the two 

 Nova Scotia genes. 



Further experiments were made with females 

 (obtained by crossing over) in which only the ^4eft half '^ 

 of a Nova Scotia chromosome was present (third line), 

 the other half being derived from an ordinary chromo- 

 some. The offspring of such a female showed that cross- 

 ing over was decreased only in the left half. 



When the right half of the Nova Scotia chromosome 

 was present (fourth line) that half was ' ^ shortened. ' ^ It 

 follows that there are two (or possibly more) factors 

 present, one in each half of the second chromosome of the 

 Nova Scotia stock, each inhibiting almost completely 

 crossing over in its own region, but not in the other region. 



An equally surprising result was obtained from a 

 female so constituted that the right halves of both mem- 

 bers of this pair of second chromosomes were present, ie., 

 when she was homozygous for the ^' right hand" pair of 

 factors for little crossing over. Under these circum- 

 stances, the crossing over was normal for this end (last 

 two lines). How such results are produced (quite aside 

 from the nature of the factor producing them) is unknown. 

 Almost inevitably, however, we think of the cause as a 

 difference in the length or shape of the chromosome con- 

 taining these factors, so that corresponding levels do not 

 come together, hence failure of interchange. When, how- 

 ever, both chromosomes are affected in the same way their 

 corresponding regions might be expected to come to- 

 gether and cross over. 



The preceding results of Sturtevant^s suggest the 

 possibility that all genes may have an effect on crossing 

 over — possibly one might think that in some mysterious 

 way the crossing-over values shown by the genes are a 



