438 ON THE IMPROVEMENT OF PLANTS. 



nurseries at Waltham Cross. The flowers of some of 

 these self-sown plants are much larger than others, 

 some are of a deeper colour, and some of a smoother 

 aspect and more rounded form; the habit of growth of 

 the different seedlings also varies greatly. The Dog-Rose 

 (Rosa canina) is another case in point. Examples of this 

 fact are, however, plentiful enough in Nature, and might 

 be adduced almost without limit if required. But it is 

 only by selection and cultivation that the most attractive 

 forms of these natural variations can be fixed and turned 

 to practical account. The advanced Lychnis, if neglected 

 and allowed to seed in its native wilds, would probably 

 produce offspring for the most part similar or inferior to 

 itself except rarely and at long intervals of time. If, 

 however, removed to a superior soil, more highly nurtured, 

 and the seeds gathered from the most attractive plants 

 only, the improvement would probably go on gradually 

 but regularly from generation to generation. 



To elucidate further the principle and results of 

 selection, let us suppose an intelligent agriculturist ob- 

 serving in his wheat-fields some individual plant of wheat 

 more vigorous in growth, or more prolific, longer, stouter, 

 and fuller in the ear than those by which it is surrounded. 

 He wishes to retain this variation, knowing that if he can 

 succeed in so doing he will thereby increase his crop. He 

 sets a mark upon this plant, saves and sows the seeds 

 separately, selects again and again, from year to year, 

 those most in advance, be they few or many, until the 

 variation first noticed has become constant and fixed, and 

 perhaps even further developed. This process of selection 

 usually requires to be pursued through several generations, 

 some of the individuals from each successive sowing con- 

 tinuing to advance on the line first indicated, and the 

 whole drawing closer together, till in the end he establishes 

 the purity or constancy of the race. Once established or 

 fixed the variation will remain an improved race or breed, 

 and if the individual plants composing it are not absolutely 



