CROSSINGOVER IN DROSOPHILA 189 



Discussion of data 



Synaptic stage probable point of crossingover. In their Mechan- 

 ism of Mendelian Heredity, Morgan, Sturtevant, Muller and 

 Bridges give on page 131 a full discussion of the possible stages 

 in oogenesis at which crossingover may occur. These authors 

 show that an interchange between homologous chromosomes 

 may take place either at an early stage in the growth period when 

 the chromosomes are fine threads, or at a late stage when they 

 are thicker and shorter. The second point was the one taken 

 by Janssens as the basis for his chiasmatype theory. The first 

 named authors do not suggest which stage appears to be the more 

 probable point of crossingover. Later Muller ('16) concluded 

 that the fine thread stage appeared to be the more likely point, 

 and recently Bridges has inclined to the same conclusion. By 

 analogy with other forms the stage which formed the basis of 

 Janssens' chiasmatype hypothesis — the diakinesis, as it is often 

 called — must occur after the egg has completed its growth and 

 is practically ready for fertilization, which undoubtedly precedes 

 polar body formation. If the crossingover occurred at this point 

 any change caused by temperature would become evident at 

 once (or, allowing the same 'incubation period/ after about 100 

 eggs had been laid) in the percentages of crossingover shown by 

 the offspring, for there are not more than ten eggs in the two 

 ovaries which are in this stage or beyond. This point taken 

 in conjunction with the positive evidence from the counts 

 given in this paper makes it highly probable that crossingover 

 occurs in the very early growth period, when the chromosomes 

 are known to be in the form of thin threads, and at that point 

 only. 



Disproof of reduplication hypothesis. A word must be said 

 regarding the reduplication hypothesis. This interpretation of 

 linkage by Bateson and his followers involved a set of divisions 

 at which certain factors were segregated, resulting in a set of 

 cells which developed at unequal rates. Sturtevant has already 

 shown that the theory is a mathematical absurdity when applied 

 to a set of linked factors in Drosophila. It is evident that the 

 experimental data given in this paper absolutely disprove the 



