vii NATURAL CROSSING 117 



Prevention. The evidence as to the means by which 

 the cross-pollination takes place is not yet as full as we 

 should like it to be, but the greater part is performed by 

 bees. Tests for wind-blown pollen made by the author 

 with glycerine smears on glass plates have given negative 

 results on the breeding plot, though a certain amount of 

 pollen must be dislodged in this way in close-sown field 



FIG. 51. NETTED PLANTS. 

 First crosses and parents. 



crop.* The remedy, therefore, seems to lie in the 

 exclusion of bees from the flower. In most countries this 

 can be done by covering the flowers with paper bags, but 

 the method fails in Egypt, since about 95 per cent, of 

 the flowers thus treated are promptly shed ; this shedding 

 appears to be due to the local interference with transpira- 

 tion, and consequent over-heating of the tissues. We 

 have therefore employed mosquito-nets, which cover the 

 whole plant, being supported over it on four posts. 

 (Fig. 51). Practically no vicinism then takes place, though 



* See Allard, H. A. (2). 



