360 COTTON 



Third year in laboratory. 



(1) Repeat tests of first and second years. 



(2) Submit 7 Ib. samples fo<r brokers' report through 

 the Imperial Institute. 



Fourth year in field. 



Plant 10 to 20 acre blocks with the finest selections, 

 and when harvested compare yield and calculate com- 

 parative return on crop valuation, keeping as the founda- 

 tion for all future selection and seed distribution the 

 progeny of the two most profitable families. 



In conclusion, it is recommended that the experimenter 

 use greater care in making his initial selections, as the 

 highest standard of perfection can only be attained by 

 careful work and multiplication from the individual plant; 

 but in order to use a hard hand in roguing it is well to 

 start with not less than 500 carefully selected plants, as 

 they rapidly decrease in the first two years. 



The system of selection discussed in this paper is no 

 doubt open to many theoretical objections, but it is 

 sufficient for the writer that it can be largely employed 

 at little expense by the intelligent planter, and even in a 

 modified form has given most excellent results in Nyasa- 

 land. 



It may be of interest to mention that the whole of the 

 native cotton in the Mlanje District of Nyasaland 

 during the current year is the progeny of two plants 

 first selected in 1909, and multiplied on the Govern- 

 ment Farm, Namiwawa, to the extent of 160 acres, and 

 then further multiplied by the villagers of two native 

 chiefs, the cotton being purchased by the British Cotton 

 Growing Association and the seed kept separate in 

 sufficient quantity to stock this district, which in a normal 

 season produces 200 tons of cotton. 



The value of a working system of seed selection in 

 cotton has been amply demonstrated in the improved yield 

 and prices obtained in Nyasaland, and the continuance of 

 a Government seed farm is a necessary adjunct to the 

 native cotton industrv. 



