No. 562] SHORTER ARTICLES AND REPORTS 



635 



If this is true, we are confronted with the problem of explain- 

 ing the apparently universal occurrence of self-red cobs in con- 

 nection with self-red ears arising in F x from variegated-eared 

 parents. Why, in short, should V c and V p , if they are really 

 distinct, always change together to S c and S p , whenever either 

 one changes ? This seems the more unaccountable when consid- 

 ered in connection with the fact that the change often, or per- 

 haps always, affects only one of the two like (duplex) factors 

 of a homozygous somatic cell, so that V C V P ■ V C V P becomes 



In my former paper (loc. cit.) I accounted for perfect 

 coupling of cob and pericarp factors in certain crosses by the 

 assumption that the two factors were located in the same chro- 

 mosome, and explained perfect allelomorphism of the same fac- 

 tors in other crosses by the assumption that the two factors were 

 located in homologous chromosomes. This was on the further 

 assumption that homologous chromosomes separate at the reduc- 

 tion division exactly at the plane of their union in synapsis. 

 If in place of this last assumption, however, we accept Morgan 's 4 

 suggestion, based upon cytological evidence presented by Jans- 

 sens, that homologous chromosomes may become spirally twisted 

 together in synapsis and that the plane of separation may not 

 always coincide exactly with the plane of union, we must also 

 accept his further suggestion that the linear position of factors 

 within a chromosome has much to do with the degree of coupling 

 and allelomorphism, "linkage." To me Morgan's hypothesis 

 seems the most reasonable interpretation of the facts of partial 

 coupling and "repulsion," and it also affords a satisfactory ex- 

 planation of perfect coupling and allelomorphism. 



In accordance with Morgan's hypothesis, we must suppose, 

 not only that the factors V c and V p are located in the same chro- 

 mosome as I had done before, but in addition that they are situ- 

 ated very close together in this chromosome, since their linkage 

 seems to be perfect. Similarly we must suppose, not only that 

 V P and S p are in homologous chromosomes, as I had previously 

 done, but that they are in almost exactly homologous positions 

 in these chromosomes, since their allelomorphism appears to be 

 perfect. This second supposition follows of course as a corol- 

 lary of the first one if 8 is produced through a modification of V. 



Now we might suppose further that the two factors, V p and 



* Science, X. S., 34: 384, 191 L 



