THE ROGUE WALLFLOWER 269 



N (? X r ? 



Ni- j, 

 (hybrid) F, 



Seed may be produced by the Nr plants either by intercrossing 

 among themselves or by crossing from the normals, and the results 

 will be different in these two cases : 



1. Where the hybrids intercross : 



Nr X Nr 



NNf Nr| Nrf rr $ 



(normal) (hybrid) (rogue) 



and rogues will reappear in the proportion of one to three, two of 

 which are hybrids. 



2. Where the hybrids cross with the normal (the most likely 

 thing to happen) : 



NN <? X Nr { 



NN f Nr f 



(normal) (hybrid) 



normals and hybrids being produced in equal proportions. Besides 

 these there are numerous normals breeding true, so that the 

 proportion of hybrids to normals will diminish in succeeding 

 generations. They could be eliminated by breeding from indivi- 

 duals, but as this is not done the hybrid type persists and there is 

 always the possibility (although a remote one) of a rogue being 

 thrown off, even without the possible occasional variation that is 

 usually supposed to account for the production of these monstrous 

 forms, and quite apart from the physical environment of the 

 plants grown for seed. 



AZOLLA CAROLINIANA Willd. 



By W. H. Burrell, F.L.S. 



There is good evidence to show that Azolla filiculoides Lam. 

 has established itself in the Thames Valley and elsewhere, but it 

 is desirable that botanists should understand the obscure status 

 of A. caroUniana Willd. in the list of British plants. The value 

 of the only vegetative character by which it was supposed the 

 two species could be distinguished when barren has been ques- 

 tioned — " The character of the hairs of the leaves does not seem 

 to be decisive " (Bot. Ex. Club Eep. for 1913, p. 31G) and we 

 have now the statement that no fruiting material has been found 

 in the British Isles (Marsh, Journ. Bot. 1914, p. 212). There 



