KDIÏII 11. S\I\I)KIIS. THF, l!l!i:KI»IN(i (H- Uni Itl.F Fl.llW MIS. 



THE BREEDING OF DOUBLE FLOWERS' 



l!y Edith R. SAUNDERS 



LecliiÈ-fraiiil lalf Kellow, Nruiiliam Cull.'ge. Caiiilii lIrc (KiiRL-ind). 



The ocriincnco of " doiibling " in llowcrs is a phenomonon willi wliich wc 

 are now faiiiilinr in an inei'casingly large number of cnllivaled planis. Tlic 

 original cause ol" lliis «loiibling, and Ihe précise nalure di' liie modilicalion 

 which tiie flo\V(M' lias undergone in beconiing double, in eacli particular case, 

 are consideralions whicli lie oulside Ihe scope of tlie presont paper. In wliaL 

 foUows I propose to deal rather with Ihe praclical question of Ihe lirerdinu of 

 double-llowered planis. 



\\'e may dislingnish Ihreo types or grades of doiibliiig : 



I) In Ihe extrême lypc Ihe Oower becoines so double Ihal neilher pollen 

 nor ovules are produced as e.g. in Arabis, Cardaniine, and the old-fashioned 

 double English Walltlower {Cheiranthus) — ail niembers of Ihe N.O. Cruci- 

 ferae. The propagation of doubles in cases of this class can, as a luIc, only be 

 effecled by végétative methods, from doubles. 



!2) In Ihe second class the organs of one sex on/;/ hecoinc ahoricd wliile 

 those of the other are still able to function. In some jilanls it is llic luale 

 organs which cease to function, in olhers the feniale. 



a) The double Pétunia alfords an instance of a case where Ihe feniale 

 ovgann are aborted, bul where good pollen is produced in large quantities, the 

 number of anthers being generally many more than in the single llower. The 

 female organ in this case is generally présent, but is more or less malformed so 

 that pollination gives no resuit. In (he more monstrous forms Ihe ovules are 

 wanting allogetlier and the ovary may be filled wilh ninnerous petaloid struc- 

 tures and anthers. 0( récent years seed of seed-producing doubles has occa- 

 sionally been listed in seed catalogues, but noue of this seed which I hâve yet 

 been able lo procure has proved to be good. It may Ije that where good seed 

 is oblained from a plant apparently double, the individuel will be found on fur- 

 ther examination to be a single, owing ils doul)le appearance lo Ihe large de- 

 velopment of the petaloid prolongations of the connective of the slamens — 

 prolongations which, as small lanceolate structures net appearingabove the top 

 of the corolla lulje, are not uncommon in many cullivalcd forms of the single, 

 but which may occasionally reach larger dimensions. 



b) In the genus Diinilhus, on the other hand, the double flowers may be des- 

 litule of pollen while the female organ reinainx normal. This is not infrequently 

 the case in the ('arnation (£*. Cari/oplii/ll/iA) and is very generally so in llie Swecl 

 William (D. barhaltis). 



7>) In the remaining and by far the largest class, doubling occurs wilhout 

 involving loss of function in the reproductive organs of either sex, as e.g. in 

 Ihe double form of Wallflower (C/ieiranlltus) more recently introduced by Ger- 

 man growers, in Ihe lloljyhock [Althaea]. and many other gênera. In doubles 



1. l'.iiiiiiiMiincalii'ii lalli' à l.a Iniisic'me sùanoe i.k' la iiiiiréii'iirr. 



