XVI] Practical Application 297 



these together, some will be impure and throw the reds 

 (Plate VI, figs. 7, 13, 19), and he may go on for some 

 years saving collectively from the dark magentas but still 

 find a proportion of reds produced. The first time he saves 

 from each individual separately he will find that so77ze only 

 are impure and others pure. Then by saving from the pure 

 plants, his pure strain is immediately established. 



When the desired combination has appeared, the readiest 

 way to perpetuate it is by self-fertilisation. But if, as 

 happens more often than is commonly supposed, there is 

 any considerable degree of self-sterility, the plant must be 

 fertilised by some other individual of the same type if that 

 can be obtained, or by one which is suspected of being a 

 heterozygous form containing the selected variety. 



The advice to attend very carefully to this matter of 

 self-fertilisation may seem superfluous, but I know actual 

 cases where practical men have attempted for several years 

 to fix a new variety of a plant growing in the open ground 

 by merely leaving the individuals uncovered, exposed to the 

 visits of insects, though in an adjacent bed were plants of 

 the original type from which the novelty had been derived. 

 Year by year the proportion of plants which came true to 

 the new variety continued very small, and the fact was 

 accepted as a symptom of especial difficulty in fixing that 

 particular variety. A few yards of muslin arranged as a 

 cage over the plants and a few minutes spent in pollinating 

 the covered flowers would have saved much further trouble, 

 and the variety could have been raised true or " fixed " in 

 one season. In reality, of course, the supposed "tendency 

 to throw back " to the parent type was due simply to the 

 pollination of the variety by insects which had visited the 

 adjacent dominants. 



The animal breeder, as he cannot self-fertilise his pro- 

 ductions, must follow a rather more complicated procedure, 

 but by the use of Mendelian methods he also can work with 

 certainty. He must take his birds or other animals and 

 test them for purity individually, usually by breeding each 

 first with a recessive, and then having found a pure 

 individual of each sex, he can by breeding these two together 

 create material for building up a pure strain. 



