xvi] Practical Application . 301 



conditions are continually being crossed together by insect 

 agency. To purify such crops by " rogueing " is an in- 

 terminable and hopeless task. The proper course is to 

 identify the factors governing the various characters, and 

 having ascertained their relations to each other, to build up 

 a desirable strain by individual selection. In former times 

 the confusion of types in the crop would have passed for 

 " variability." Mendelian analysis shows that representation 

 of the facts to be entirely mistaken. 



Technical Methods. 



The technique of Mendelian experimentation is usually 

 very simple. In crossing plants together the anthers of the 

 plant to be used as female must of course be picked out 

 with forceps before they dehisce. The other parts should 

 be injured as little as possible. The flower is then covered 

 to exclude insects. Muslin bags may be used for this 

 purpose, but they are neither so convenient nor so safe 

 as rain-proof bags made of parchment-paper, which most 

 manufacturing stationers can now supply in any required 

 size. The bag is put over the flower, the mouth being 

 crushed up so as to fit to the stem, and it is fixed in place 

 with a small bent piece of copper wire. If the stem be 

 delicate a thin bamboo stuck in the ground must be also 

 held in the copper clip. The use of the clip obviates 

 all tying and untying. Muslin bags are objectionable for 

 various reasons, especially because unless they are con- 

 tinually readjusted, flowers are sure to touch the sides and 

 there is the risk of fertilisation by insects. Bees will often 

 visit flowers covered by muslin, and I suspect that nocturnal 

 moths may do the same. If it is desired to give ventilation, 

 holes may be punched in the top of the paper bags, above 

 the flowers. The flower from which pollen is to be taken 

 must also be covered before it opens, in order to keep its 

 pollen from pollution. There are indications that some of 

 the results obtained by the older hybridists were confused 

 by neglect of this precaution. 



In transferring pollen the use of brushes is to be 

 deprecated, as tending to the introduction of errors. The 

 best plan is to pick out with fine forceps an anther from 



