A. If. Trow • 297 



(6) no crossing over \ 



(c) single crossing over > when l,lu; V chroiuDsoiiiu is excluded. 



(d) double crossing over J 



Why should a chroinosonie behave difteri'iitly as in b, c, and d, or 

 indeed in all four cases i* Surely, the cells concerned, and their nuclei, 

 are for all jjractical purposes indistinguishable from each other. 



3. There is a logical fallacy underlying the whole of the crossing 

 over hypothesis, as applied to the location of the ftictors. Let it be 

 granted that under certain limited conditions the number of cross-overs 

 is proportional to the distance of the factors from each other. It does 

 not follow that the distance of the factors from each other is proportional 

 to the number of the cross-overs. It may be true that " all men are 

 fools " ; it does not follow from this that "all fools are men" 



In truth, the foundation on which the hypothesis of crossing over 

 has been built up appears to be very unstable. Ruptures and recom- 

 binations take place without any adequate cause. Why should a 

 chromosome break where it crosses another? If it break, why not 

 remain broken ? If it recombine, what regulates the recombination ? 

 What force secures the absence of crossing over where the Y chromo- 

 some is concerned ? What determines the constant numerical relation 

 between the no-ruptures, single ruptures, and double ruptures ? 



Finally, let it be noted that the graphic representation of the 

 location of the factors is a type of representation common to every set 

 of phenomena which can be expressed as percentages. The exponents 

 of the reduplication hypothesis may very well accept the plans of the 

 chromosomes, as graphic representations of the relative strengths of 

 reduplication. For their purposes the diagrams must be read fiom 

 the 50 °/^ mark (no reduplication) towards the two ends. 90 7,, is 

 indeed a gi'aphic representation of — 10 °/,^ (repulsion). 



In this analysis it was not thought worth while to pursue in detail 

 other but somewhat petty difficulties, such as the presence of a spherical 

 chromosome, with cross-overs associated with it ; the probability of 

 untwisting taking place in metakinesis ; the difficulty of demonstrating 

 a rupture and recombination by the existing cytological methods, and 

 so on. It may be noted, however, that Drosophila with its few chromo- 

 somes, may shew an unusual amount of chromosome isolation and of 

 reduplication (linkage and crossing over). Whatever may be the fate 

 of their ingenious but, I fear, rather overworked hypothesis, students 

 of genetics will continue to welcome the detailed contributions of these 

 American authors. 



