Volume II 



FEBEUAEY, 1913 



No. 4 



FOEMS OF EEDUPLICATION :— PEIMAEY 

 AND SECONDAEY. 



By a. H. trow, D.Sc, F.L.S. 



According to Bateson and Punnett (/. of Gen. Vol. i. 4, p. 293) all 

 forms of reduplication Initherto discovered, with one possible exception, 

 may be grouped in two classes, the gametic series for these being 

 represented by the empirical formulae 



n—\:\:\:n—\ and 1 : « — 1 : ?i — 1 : 1 ; 



the single exception being the cases of complete repulsion, in which the 

 end terms disappear, giving the series n — \ :n — \. Even this exceptional 

 type however is now regarded by these authors as probably a special 

 case of the series 1 : m — 1 -.n—X : 1, where n is large and therefore one 

 of the zygotic types so scarce as not to be expected except in very 

 extensive cultures^ 



My own studies of Senecio vulgaris have however revealed the 

 existence of the ratio 2:1:1:2, and Baur (Vererbungslehre, p. 124) 

 appears to have found the ratio 6 : 1 : 1 : 6 in an Antirrhinum cross. 

 Neither of these ratios comes under the general formula given above. 

 Such being the case, it seemed to me desirable to ascertain what the 

 consequences of accepting the current hypothesis of reduplication 

 would be, not simply as applied to a pair of factors AB (or two pairs of 

 allelomorphs Aa, Bb), but to three or more factors A, B, C, D .... 



The immediate problem which presented itself for solution was 

 a comparatively simple one, yet one which has apparently been over- 

 looked. Given three factors A, B, and C and the occurrence of re- 

 duplication between A and B in the form 7i : 1 : 1 : w and between A and 

 C in the form m : 1 : 1 : m, where n may be equal to, greater, or less 

 than m, is there necessarily a form of reduplication between B and C, 



1 In these formulae n is a power of 2 and is equal to one-half the number of gametes 

 in a series. 



Joum. of Gen. n 22 



' i.W YOKK 



HUlAINICAL 



UaKUI£N. 



