32 " Rogtfes" in CnluKir// Peas 



plants are those ivhich would take part in self-fertilisation, and that we 

 are not concerned with any phenomenon of selective fertilisation. The 

 results of the crosses must therefore be in reality heterozygous and 

 must receive fi-om the one type-parent those elements which if they 

 were united with a similarly constituted gamete of the other sex would 

 form a type-plant. Nevertheless such plants are rogues and breed 

 rogues only. Therefore the type-elements received by the F^ from the 

 type-parent must be permanently lost. Whither do they disapi)ear ? 

 The onlij answer to this question which we can offer is that when intro- 

 duced from one side only of the jxirentage these elements are in some 

 way used ujJ and cut out of the germ-lineacje in the early stages of the 

 somatic development. The young stages of the F^ plants are, almost 

 always, type-like, but the characters of the type are left behind with 

 further development. 



We have also evidence that a similar change takes place in the life- 

 history of some at least (probably most) of the rogues which come as 

 the immediate offspring of types. Several times families afterwards 

 found to contain rogues, though individually examined in the early 

 stages, w«re then recorded as free from rogues ; and occasionally in 

 these families plants which developed into ordinary rogues are recorded 

 in the juvenile condition as doubtful. Rogues, offspring of rogues, 

 never pass through such a stage. There is therefore no reasonable 

 doubt that the rogues arising as the offspring of types are at all events 

 fi-equently heterozygotes formed by the union of type and rogue gametes, 

 and since they always breed true, in them also the type-elements must 

 be lost in some somatic stage. 



Some light may perhaps be obtained by examining the produce of 

 seeds from the various parts of the plant separately and this we shall 

 endeavour to do, but it is likely that by the flowering stage the 

 differentiation, or segregation if the term be applicable, has already 

 been effected. Nevertheless in plants of the intermediate class the 

 apical parts are so much more rogue-like than the rest that the experi- 

 ment ought to be made. 



We incline to think these indications point to some process of 

 somatic segregation which prevents the type-elements from reaching 

 the germ-cells of the cross-bred plant. A comparison may perhaps be 

 suggested between this phenomenon and the examples, fairly frequent 

 in horticultural literature, of plants which produce from root-cuttings 

 forms different from those arising by the propagation of shoots or buds 

 above the root-system. 



