406 SELECTING AND IMPROVING PLANTS IN CULTURE. 



may be of more diminutive size, or its shoots may take a different direction, 

 as in fastigiate and pendulous-branched trees. All these, and many other 

 accidental variations, which, as we have seen (551), cannot generally be 

 reproduced from seed, may be perpetuated by cuttings, or some other mode 

 of propagating by division. 



865. Cross-breeding. This process is effected by fecundating the stigma 

 of a flower of one plant with the pollen from the flower of another of the 

 same species, but of a different variety. Sometimes fecundation may be 

 effected with the pollen of a different species, and in that case the produce is 

 said to be a hybrid, while in the other the result is merely a cross or a cross- 

 bred variety. The mode of performing this operation has been very well 

 described by Mr. Hay ward. " Supposing," he says, "for instance, you have 

 two geraniums producing differently-shaped leaves and differently-coloured 

 blossoms or two apple-trees, bearing apples of different sizes, colours, and 

 qualities, and it be desired to produce geraniums of differently-shaped leaves 

 and differently- coloured flowers, and apples of different sizes, colours, and 

 qualities, that is, different from either of the two plants or trees which you 

 possess : the mode of effecting this is to select a blossom of the plant from 

 which you wish to obtain the seed ; when it is just on the point of opening and 

 exposing the anthers, take a pair of scissors, and, gently forcing open the petals 

 of the blossom intended to bear the seed, cut off the stamens, and remove the 

 anthers, and then leave the blossom thus operated upon for a day or two, or 

 until the petals are quite expanded, and the pistil arrived at a state of 

 maturity ; when it is in this state, select a blossom of the plant with which 

 it is desired to impregnate the prepared female blossom, and when this is in 

 a state of maturity, and in a state to part with its pollen or farina freely, 

 take a small camel's-hair pencil, collect the farina on the point, and place it 

 on the stigma or crown of the pistil of the prepared blossom. This opera- 

 tion may be performed, with an equal chance of success, on plants of all 

 descriptions." (An Inquiry, c. p. 120.) " New varieties of every species 

 of fruit," Mr. Knight observes, " will generally be better obtained by intro- 

 ducing the farina of one variety of fruit into the blossom of another, than by 

 propagating from any single kind. When an experiment of this kind is made 

 between varieties of different size and character, the farina of the smaller 

 kind should be introduced into the blossoms of the larger, for, under these 

 circumstances, I have generally (but with some exceptions) observed in the 

 new fruit a prevalence of the character of the female parent ; probably 

 owing to the following causes. The seed-coats are generated wholly by the 

 female parent, and these regulate the bulk of the lobes and plantule : and I 

 have observed, in raising new varieties of the peach, that when one stone 

 contained two seeds, the plants these afforded were inferior to others. The 

 largest seeds, obtained from the finest fruit and from that which ripens most 

 perfectly and most early, should always be selected. It is scarcely necessary 

 to inform the experienced gardener that it will be necessary to extract the sta- 

 mina of the blossoms from which he proposes to propagate, some days before 

 the farina begins to shed, when he proposes to generate new varieties in the 

 manner I have recommended." (Knight's Physiological Papers, p. 177.) 



866. Precautions against promiscuous fecundation require to be taken 

 both in the case of flowers the seeds of which are to be sown for the purpose 

 of selection, and in those which have been cross-fecundated. In the former 

 case, the plants should as much as possible be isolated from all others of the 



