86 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



corroborate, quite as often modify or cause the abandonment of 

 the conclusions of the past. Some few years hence, Deo volente, 

 perhaps the Society will allow me the attempt to come before it 

 with a discourse fuller of compact and useful information. 



The hybridist who goes about his work thoughtfully has, I 

 take it, three chief objects in view : — 



(1) The improvement of existing varieties. 



(2) The advancement of our knowledge of the phenomena of 

 cross-fertilisation. He will not be content with the acquire- 

 ment of a " rule-of-thumb " art of producing improved plants, 

 but by systematising and recording his methods will help on- 

 wards the endeavour to reduce cross-fertilisation to an orderly 

 science. 



(3) The verification or correction, under the light of his own 

 experiments, of the work of his predecessors in the same field ; 

 also the examination of the work of Nature. The raisers of 

 cross-bred Orchids have again and again reproduced artificially, 

 and thus assigned with certainty to their parentage, natural 

 hybrids whose origin was before unknown or merely guessed at. 



When, a good many years ago, I sowed my first Narcissus 

 seed, it needed some boldness to aim at excelling the beautiful 

 flowers bequeathed to us by former workers, and, from the 

 florist's point of view, the sentiment " pereant qui ante nos 

 nostra fecerunt " was almost excusable. But from the scientific 

 side it was evident that much remained to be accomplished. 

 Messrs. Leeds and Backhouse, from whom, with a few notable 

 exceptions, we inherit this wealth of "Daffodils," which 

 has changed the face of our gardens and markets in spring, 

 have left a very imperfect record of their methods in detail. 

 Leeds, indeed, the author of half these flowers, has given us no 

 clear information to speak of, and Backhouse, who raised the 

 finest, in the short account contributed to the Gardeners' 

 Chronicle of June 10, 1865, seems to give a somewhat vague 

 summary from memory rather than to draw from accurately kept 

 notes. Thus he does not specify the parentage of his remarkable 

 productions "Emperor" and "Empress" more particularly 

 than to say they were the offspring of N. bicolor, a name which 

 itself denotes more than cne form, and some other variety of 

 N. pseudo-Narcissus. Indeed, although they have been produced 

 within living memory, the precise origin even of whole classes or 



