4 BULLETIX 644, U. S. DEPARTMENT OF AGRICULTURE. 



The lint index given for a seed weight of 13 + 25 per cent is practi- 

 cally identical with that given for a seed weight of 6 -f- 42 per cent. 

 In each case practically the same quantity of fiber is obtainable from 

 a given number of seeds, although one variety has a lint percentage 

 of 25 and the other 42. Thus it is seen that the larger percentage of 

 lint is due entirely to a decrease in the weight of the seeds without 

 change in the amount of fiber per seed, a striking illustration of the 

 fallacy of basing an opinion as to the value of a variety of cotton on 

 the percentage of lint alone. 



A high percentage of lint, therefore, does not necessarily mean an 

 abundance cf fiber. The fiber is actually less abundant when there is 

 42 per cent of lint in a variety the seeds of which weigh 8 grams per 

 hundred than when there is 30 per cent of lint in a variety the seeds 

 of which weigh 14 grams per hundred. 



That the percentage of lint will steadily increase as the size of the 

 seed decreases without altering materially the actual amount of fiber 

 obtainable may be seen if the lint index under a seed weight of 

 13 -|- 25 per cent be taken as a base. Approximately the same lint 

 index may be traced diagonally across the table to the lint index 

 under a seed weight of 6 -4- 42 per cent. It can readily be appreciated 

 that this trend represents the possible results of selection based on 

 lint percentages alone. 



Conversely, an increase in the weight of the seed may reduce the 

 percentage of lint without reducing the actual amount of lint; but 

 unless the reduction in the percentage of lint is proportionate with 

 the increase in the size of the seed the abundance of the lint is also 

 increased, notwithstanding the reduction of the lint percentage. 



This is in accord with the results obtained by Mr. T. H. Kearney 

 in his work in the acclimatization of Egyptian cotton, 1 in which he 

 noted that the lint percentages of his best selections were steadily 

 decreasing from year to year below that of the original imported 

 strain, but on ascertaining the lint indexes of the selections and 

 imported stock he found that there was no actual diminution in the 

 quantity of fiber produced. The decrease in the lint percentage was 

 due entirely to an increase in the weight of the seeds. Mr. Kearney 

 concludes : 



The negative correlation between the characters lint percentage and weight 

 of seeds is sufficiently pronounced to indicate that a high percentage of lint 

 is in large measure associated with low weight of seeds. * * * It might 

 be inferred from these facts that lint percentage can be used with greater 

 safety as an index of productiveness in comparing individual plants of a fairly 

 uniform variety than in comparing different varieties. 



That the percentage of lint of a selection or variety of cotton 

 should be considered only in the light of the lint index is well illus- 



1 Kearney. T. H. Lint index and lint percentage in cotton breeding. In Ann. llpt. 

 Amor. Breeders' Assn., v. 7/8, p. 25-29. 1912. 



