512 JOUKNAL OF THE DEPARTMENT OF AGRICULTURE. 



pound. It is thus clear that length of staple is a very important 

 factor. These prices were for abnormal times. In normal times the 

 differences will not be so great, but a bigger premium will always be 

 paid for the longer staple. The sta])le must also be strong, or else 

 it is of very ]itll(.' use io the spinner. 



(2) Productiveness. — In order io get a good \icl(l, seed musl lie 

 selected from only well-formed plants, witli plenty of fruit limbs 

 and prodiiciug a large number of well-developed bolls. Plants which 

 are sterile or partially sterile should not be selected, because if they 

 are selected a decreased yield will result. 



(3) High Percentage of Lint. — If the seeds are well covered with 

 lint, we can expect to find a high percentage of lint to seed. Partially 

 naked and naked seed will give a smaller percentage of lint to seed, 



(4) Large Bolls. — Varieties with big bolls have a great advantage 

 in reducing the cost of picking. They are usually storm-resistant. 

 In a big boll variety we would nol expect to find plants maturing 

 small bolls, and should such plants ai)pear they must be taken out. 



(5) Storm Resistavce. — Storm resistance is indicated by the 

 drooping position of the bolls and also by a heavy involucre, which 

 will assist in protecting the bolls from wind and the beating of rain. 



(6) Early Maturity. — Early maturity is again indicated by the 

 position of the fruit limbs on the i)lants. Such plants are usually 

 small, with the lower limbs very close to the ground. 



(7) Uniformity. — It is very essential to have a uniform field of 

 cotton. This can ])e obtained by selection. TTniformity must be 

 maintained, for degenerate plants will always appear. A uniform 

 field of cotton will give a larger yield, and lint from such a field 

 will be of better quality and fetch higher prices than from a field 

 not so uniform. 



(8) Disease-re.^isting Qualitie.^. — Very often a number of healthy 

 ])lants can be seen in a diseased cotton field. By selecting seed from 

 such plants a strain can be develo])ed which will l)e able to resist that 

 ])articular disease. An example of this is th(^ wilt-resisting strain 

 of cotton developed in the Southern Stales. 



In selecting for the above we must not sacrifice the desirable 

 characters of the variety, but musi select for these chaiacters only as 

 far as is consistent with the variety . 



To what extent the grower will be successful in his selection work 

 will depend entirely on his skill to detect minute differences, enabling 

 him to grow a variety which will shoAv the most uniform expression 

 of desired characters. The following sub-divisions will assist the 

 grower in making ids selections: (1) Shape and structure of bushes 

 and leaves; (2) shape and size of bolls; (3) shape and colour of seed; 

 length, strength, and silkiness of fibre; (4) colour and .shape of 

 flowers. 



The first step of the cotton grower must be to decide upon his 

 variety, and then to procure his seed from some reliable source. As 

 soon as the plants are from 2 to 3 feet in height selection must begin. 

 In making his selections he must examine and study the best plants 

 in the field. In (U-der to get the best impression of the true type it 

 is desirable to look about until several of the most likely plants are 

 found together. In comparing the different plants in the same field 



