GENETICS PURPLE EYE COLOR DROSOPHILA 301 



the amount of crossing over is probably due to a change in the 

 physical properties of the chromosome substance, has an im- 

 portant bearing on the question of the stage at which crossing 

 over itself occurs, as follows: From a study of the time taken 

 for the effects of exposure to abnormal temperature to become 

 manifest or to disappear, Plough concluded that the effect was 

 produced at one stage only in the development of the ovary, 

 and that eggs which have not arrived at or have passed this 

 critical stage are incapable of registering any temperature vari- 

 ation. It was next argued that this critical stage is that at 

 which crossing over itself normally occurs. It is certain that 

 crossing over does not take place before this critical stage is 

 reached, but it does not follow that it might not occur at some 

 stage between this and the maturation divisions; that is, at any 

 later stage during the growth period. At the critical stage one 

 of the factors which modifies the frequency of the crossing over 

 becomes fixed, but the crossing over itself may occur later. As 

 a crude analogy, the process of crossing over might be likened 

 to a machine — say, a sawmill. The rate at which boards are 

 sawn depends, other factors remaining constant, upon the 

 toughness of the log fed against the saws, which toughness is a 

 physical property of the log fixed long previously. 



The coincidence analysis indicates that the setting of the 

 crossing-over machine has not been altered, but that the chromo- 

 some at a specific sensitive stage in its fabrication has been 

 modified in one of its properties — ^its toughness, let us say— so 

 that when it ultimately undergoes crossing over the output is 

 different. 



It is quite possible that the crossing over follows immediately 

 after the determination of this property; indeed, from other 

 lines of evidence it seems probable that crossing over occurs at 

 a thin-thread stage, or at least that the characteristic trans- 

 junction is accomplished at a leptotene stage, such as occurs 

 only in the early growth period. But such a conception does 

 not exclude the possibility that the crossing over occurs at a 

 four-strand stage, as is indicated by still other lines of evidence. 

 To call such a 'thin-thread, four-strand, early-growth stage,' 



