116 THE COTTON PLANT IN EGYPT CHAP. 



When n is infinite (or practically in our case when it 

 exceeds ten) : 



i rr 



and H 



x + y 



In the 75 cases already mentioned, we found, dealing 

 with two pairs of characters, that there were twenty-six 

 cases of recessives splitting out from dominants. Thus 

 ff=^% after n years 



With two pairs of characters 



2 2 1 



y = T2 or - 



4 4 



And since 



H _ x _26 



aj + y 75 

 .'. # = 0132. 



Thus there was 13 '2 per cent, of natural crossing as the 

 mean value for past years. 



By direct observation we found five vicinists from the 

 previous year, implying about ten altogether, or about 

 13 '3 per cent, of natural crossing in 1904. 



The two results agree. 



The absence of facilities has prevented the author from 

 carrying this analysis further by the use of pure strains 

 planted under field conditions, but a value of 5 per cent, 

 to 10 per cent, for natural crossing under field conditions 

 in Egypt has been confirmed by numerous, though non- 

 systematic, pieces of evidence. 



This value is expressed in terms of flowers crossed to 

 total flowers ripening. The possibility of mixed pollina- 

 tion should not be disregarded, since we shall see that 

 hybrid and selfed embryos may be formed side by side in 

 the same ovary. The expression of the value in terms of 

 ovules crossed to ovules ripening would therefore be 

 preferable. 



