20 THE DEVELOPMENT OF RAW COTTON 



had remained unchanged for nine years, the other for 

 seven ; one would not produce half a crop in the north of 

 Egypt, the other would not produce half a crop in the 

 neighbourhood of Cairo ; one consistently contained 

 about 30 per cent, more salt in its cell sap than the other 

 when the roots were occupying the same soil. Such 

 simple differences as J inch in lint length, 1 per cent, in 

 ginning out-turn, and 2 or 3 degrees in the angle of the 

 lobing of the leaf; colour of leaves and flower, shape and 

 dimensions of the boll, fuzziness of the seed, and so forth, 

 while all capable of being distorted from the normal, 

 were, it need hardly be added, all showing exactly the 

 same under the same conditions in 1913 as they had been 

 in 1907. Left exposed to crossing and mixture for a 

 single season the strains showed 10 per cent, of impurity 

 in the following year. 



Such strains as those mentioned, in conjunction with 

 other experimental evidence, throw light on such cus- 

 toms as " change of seed," " selection for 

 Interpreta- 

 tion of yield," etc. From a sowing mixture of these 



Popular two strains in equal parts we harvested 75 

 Expressions. per cent of Qne Qr the ot h erj according as to 



whether the mixture was grown in the south or the north. 

 In the following year the seed from this mixed sowing 

 produced a percentage of hybrid plants; selection of the 

 highest yielders in the population included nearly all these 

 hybrids, and scarcely any of the parental stocks. Thus 

 in the third year, had the seed of these plants been 

 sown, both the pure parents would have been lost entirely, 

 and a most intricate jumble of all sorts of cotton would 

 have rewarded this blindfold selection. 



