302 Practical Application [ch. 



the flower to be used as male, and with it to touch the 

 stiema of the female flower. The anther is then thrown 

 away, unless the supply of pollen is very limited. Absolute 

 cleanliness is of course most essential. The fingers and 

 forceps should be continually cleaned with spirit in order to 

 kill any pollen-grains adhering to them before proceeding 

 to the next fertilisation. In operating on plants with much 

 pollen like Oenothera, Poppies, &c., care should be taken 

 to avoid including in the bag leaves on which pollen has 

 already fallen, and many other small precautions of this kind 

 will occur to the experimenter in practice. For example, 

 if muslin bags are used, they should not be laid on the 

 greenhouse stage where the plants have been standing, for 

 obviously the fallen pollen may adhere to them and be 

 passed on to the stigmas. 



In sowing fine or light seed in pans the presser with 

 which the soil is patted down must be cleaned as each pan 

 is finished. Otherwise it is liable to carry on seed to the 

 next pan sown. 



In recording results the capsule of seed resulting from a 

 cross should have a register-number under which it is sown. 

 Each plant which comes up must then be numbered sepa- 

 rately, and its number is written as an index-number to the 

 original number of the capsule. Thus, a pod of Emily 

 Henderson Sweet Pea % x Blanche Burpee ^ may, when 

 sown in 1904, have the number 20. If seven plants come 

 up, they are numbered 20^~^ In 1905 each of these is 

 sown under a fresh number. 20^ may for instance become 

 115, 20^ becomes 116, &c. The families 115, 116, &c., will 

 each contain a great number of individual plants, and each 

 of these from which the breeder intends to save seed must 

 then receive an index-number 115^ 115^ &c., under which 

 its characters are recorded in the register. In 1906 the 

 offspring of each of these plants receives a fresh number for 

 the year. The family of 115^ may become 305, and so on. 

 Reference is thus made easy, and the history of any in- 

 dividual plant can be rapidly traced. The same system can 

 of course be easily modified so as to adapt it to the case of 

 animals where the mating must be bi-parental. 



