xiv MAKING NEW PLANTS 171 



was to be found. Some were purely fantastic, with 

 no special beauty ; others had beauties not present 

 in any one of the original lilies, but produced by 

 the process of hybridisation. 



This was the first step, the crossing. The next 

 step was, of course, the selection of the best kinds. 

 The fields of lilies had to be gone over with pains- 

 taking care. All the useless or ugly plants were 

 pulled up and burnt without mercy. Quite often 

 among these hybrids would be found one which 

 had some particular merit, perhaps colour or 

 perfume, and at the same time some undesirable 

 character, such as a stalk too weak to carry the 

 flower head, an unsymmetrical flower, and so on. 

 In these cases it was necessary to recross with 

 another lily, in the hope of getting new -hybrids 

 which had all the merits and none of the dis- 

 advantages. This recrossing had, of course, again 

 to be followed by fresh cultivation and selection. 



In this way more than a million lily bulbs were 

 grown, and the final result was that some two 

 hundred and fifty thousand distinct lily hybrids 

 were produced as the result of the experiments. 

 Only some of these were, however, useful. These 

 cases show us what the gardener can do by culti- 

 vation, crossing, and selection to improve plants. 



But it is not sufficient to produce a new 

 variety; it is necessary to produce it on such a 



