JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



M. Laftay, of Auteuil, and other growers, have told me the same ; 

 and I have myself had the same experience at Wissous. 



" In such a case this is what happens : The seedling Pimpernel is 

 examined and found to have no resemblance to its seed-parent, 

 Bengal, but certain characteristics are recognised in it belonging 

 to Provins or alba ; it is at once wrongly classed as a hybrid 

 between Pimpernel and Provins or alba. The Rose is simply an 

 individual that has suddenly reverted to one of its original types, 

 without any hybridisation whatever. 



" It also sometimes happens that one obtains specimens in which 

 the usual characteristics of different species are so confused that it is 

 not possible to determine them at all. Such, for example, are the Roses 

 named 1 Dona Sol,' 1 Amelie Guerin,' 1 Triomphe de Laffay,' 1 Davoust,' 

 ' Cornelie,' ' Egine,' &C, One cannot reasonably call them hybrids, but 

 nevertheless that is what is done by our growers, who class them 

 as hybrids of unknown origin. People who see signs of hybridi- 

 sation everywhere are led into error by the importance they attach 

 to the specific characteristics attribute^ to Roses by authors, and 

 these people think that they must regard every individual that 

 resembles any particular species as being derived from it. 



" This manner of reasoning, although false, has nevertheless a 

 logical appearance of truth, for to be consistent it is necessary to 

 choose one of two things : either to acknowledge hybridisation 

 wherever the specific characters of two Roses of different species 

 are found on the same plant, or only to recognise the small 

 number of well-known species whose characters are invariable ; and 

 this number may be reduced to three, namely : Rosa simplicifolia, 

 B. lutea, and B. centifolia, and this is just what one must do. 



" I cultivate an immense number of Tulips, Carnations, Ranun- 

 culuses, Auriculas, Dahlias, Camellias, &c., all differing among them- 

 selves as much and more than the varieties of Roses. Nobody has 

 been tempted to look for signs of hybridisation in the new seedling 

 varieties which are daily obtained, and nevertheless they give an 

 infinite variety of combinations of shape and colour. Why cannot 

 they allow in the case of Roses the same powers of nature, the 

 inexhaustible richness which it displays in the flowers we have 

 mentioned? Can it not just as well unite in one individual the 

 shape of the centifolia, the leaves of the Pimpernel, the scent of the 

 Provins Rose, just as it can produce in one Tulip the brightness, the 

 form and grace of three other varieties, without the necessity of mixed 

 fertilisation ? Is it more difficult to explain physiologically how a 

 wild Rose may become, by cultivation, a scented centifolia Rose, 

 than how a wild Pear tree, with long thorns, may develop into a 

 Doyenne du Cornice or a Duchesse d'Angouleme ? 



" Besides, if hybridisation were as common a thing as they made 

 out, why does it only take place with some species and not others ? 

 Can any gardener show me any centifolia, alba, Provins Roses, &c. 

 which are hybrids of Bosa simplicifolia, B. Eglanteria, B. lutea, or 

 B.punicea? No ! because these Roses have invariable characteristics, 

 which stamp them as species, and one can only obtain varieties of 

 them by submitting them to the most careful cultivation, and by 

 sowing their seeds, which only can produce that jonquil-yellow colour 

 which makes them contrast so strongly with all other Roses. 

 The same causes which produce varieties in other organic beings, 

 also produce the variations in Roses, and these causes are nume- 

 rous, although little known, without including hybridisation. 



" In conclusion, I think that this phenomenon is far from being 



