LiyXEAX SOCIETY OF LOXDOX. 83 



end of my address a arge number of lantern-slides'and specimens 

 showing this plainly : for the present, I must refer to the plates 

 here present, and wish to call your attention especially to the 

 plates with large flowers hanging in front of you, showing the 

 very great alteration in the form of flowers which result from 

 such a segregation. 



From my experiences with Antirrhinum, I think I am justified 

 in concludiiig that the difference claimed by de Vries for varieties 

 and species does not exist. One might yet object that this 

 segregation -were induced by the impure' condition in which 

 A. (flutinosum must necessarily be used, because it is self-sterile 

 and every individual in consequence itself a hybrid. I do not 

 think that this small degree of hybridity has "any considerable 

 effect on the enormous segregation going on, but fortunately we 

 have proof in the case of A. semperuirens, which can be had pure, 

 that segregation in this case also, after crossing with A. majus, 

 takes place in the same manner. 



It was further shown that sooner or later stable segregates 

 arise, so that I think the proof is there that species can arise by 

 crossing, and I personally think tiiat new species always originate 

 in consequence of a former cross. 



Now, if my contention is right, that species are stable and not 

 affected in their hereditary constitution in any way except bv 

 crossing, it must be possible to re-obtain, with the certainty of a 

 chemical reaction, any desired form — provided it is not of too 

 rare an occurrence among the segregates, — if we can only obtain 

 the pure species from which this forui was originally obtained. 

 I have tried to put this contention to tlie test ! 

 Shortly after the introduction of Petunia violacea, about 1S30, 

 this species was crossed with P. nj/ctaginijfora, and soon after 

 Petunias were obtained with green-rimmed petals. These have 

 since disappeared, as has also the original P. violacea. It has been 

 pointed out by Bailey and by Correns that the P. violacea of the 

 trade, which was used by the latter in crossing, is not the original 

 P. violacea. By a lucky chance, howe\er, this original P. violacea, 

 cultivated since its introduction, is still present in the cultures 

 of Vilmorin. AVith it, kindly given to me by my friend, Philippe 

 de Yilmorin, I have crossed P. nyctaginiflora, and among the 

 segregates in F^ obtained in large numbers the long-lost green- 

 rimmed Petunias, drawings of which are here present. 



P. violacea and P. nyctaginiflora have thus retained in Yerrieres 

 their original constitution and produce now, as in 1830, green- 

 rimmed Petunias with the certainty of a chemical reaction. 



This certainly speaks strongly for the constancy of species, but 

 I think here is a case which brings almost irrefutable evidence of 

 this. It is the case of Ciitisvs Aclami, which, as we all know, is a 

 graft hybrid between Cijtisus purpureus and Cytisus Laburnum. 

 For more than a hundred years the cells of these two different 

 species have lived in intimate connection, yet the branches arising 

 '• by reversion " furnish seeds from which both species arise in 



