30 " Rogues " in Culinari/ Peas 



2nd Exception. In this (f |) a cross of D. A. rogue x D. A. type gave 

 5 plants, of which 3 were rogues and 2 were intermediate, being the 

 only D. A. plants we have bred which look comjmrable with the definite 

 intermediates so often bred in E. G. Their offspring will be seen in 

 1915. 



We have no interpretation of these remarkable exceptions as yet 

 to offer. 



Crosses with Inter mediates. 



Since crosses between rogues and types give rogues only (with the 

 two exceptions named), it was to be expected that crosses between 

 rogues and intermediates would do the same. This cross was -made 

 only once and gave six rogues (f|). 



More interest attaches to the crosses between the intermediates 

 (class 3) and the type. The result of this mating is known in three 

 cases only (see Table, 49, 50, 69). It will be observed that in one case 

 the tyjjc E. G. was the father, and in the other two the type D. A. was 

 used as mother. The evidence i.s meagre, but it supplies proof of the 

 important fact that neither sex of the class 3 intermediates can be 

 homogeneous. 



Discussion. 



The general course of the phen(jmena is evidently quite unlike 

 anything with which we are familiar in ordinary Mendelian inheritance. 

 Since the types can throw rogues and the rogues cannot throw types, it 

 seems clear that the types contain something which the rogues do not 

 contain. This something however is different from an ordinary Men- 

 delian factor both in the effects of its presence and in the manner of 

 its distribution among the gametes. 



As stated in the introduction the visible differences (apart, that is, 

 from the difference in flavour of the seeds) between type and rogue are 

 essentially quantitative. The distinctions affect the shapes of the 

 organs, but possibly all of them, even the striking difference between 

 straight and curved pods, may be consequences of greater or less ex- 

 tension, and we incline to think that the rogue is really a form wanting 

 in some particular kind of luxuriance or jDower of amplification. The 

 difference is of course not merely one of size, for rogues may be large 

 and types may be small ; but it must be rather some quality of ex- 

 tension, dependent on cell-division occurring chiefly at right angles to 

 the axis of growth. 



