COTTON AND SORGHUM 177 



homozygous. Correlation between length of lint and size of 

 seed may explain some complications. 



Mutations in Cotton. The cotton plant, like (Enothera, has 

 often been spoken of as having germinal instability and likely 

 to produce mutations. While mutations undoubtedly do occur, 

 it is likewise highly probable that many of the so-called mutations 

 are simply segregates of a former natural cross. The ease with 

 which natural crossing occurs and the large number of chromo- 

 somes (20 according to Balls) contained in the cotton gamete 

 facilitate the appearance of forms differing from the general type. 

 The larger the number of haploid chromosomes the more difficult 

 it is to secure homozygous individuals after a cross. Egyptian 

 cotton is described by Kearney (1914) as being a mutating type. 

 From it the varieties Yuma, Pima, and Gila are supposed to 

 have arisen. In this connection it is of interest to point out that 

 the common belief as to the origin of Egyptian cotton is that it 

 arose by hybridization between a brown-linted tree cotton and 

 American sea island. The subsequent development is unknown. 



In view of the foregoing and the fact that no convincing evi- 

 dence has been presented to the contrary, the present writers 

 believe that many of these supposed mutations are in reality 

 factorial recombinations resulting from natural crossing. 



Cotton Breeding. Cotton improvement by breeding may be 

 sought along lines similar to those followed with all naturally 

 selfed crops. In producing pure-line material for scientific 

 study and subsequent hybridization it is essential to obtain ab- 

 solute self-pollination. 



From a commercial standpoint, a productive cotton with long 

 lint and smooth seed is desirable. Webber (1905) crossed 

 Klondike, a productive upland variety, and sea island, which has 

 long lint and smooth seed. Out of an F 2 generation consisting 

 of several thousand plants, only 12 combined the large blunt bolls 

 of the upland with the long lint and black seed of sea island. 

 The progeny of each of these 12 plants was grown in isolated 

 plots and subjected to vigorous selection. In the fifth generation 

 a number of plants gave progeny " nearly fixed in type." 



Resistance to wilt disease is a character of considerable com- 

 mercial importance. This disease is caused by Fusarium vasin- 

 fectum Atk. which according to Orton attacks only the cotton and 

 its near relatives. By the plant-to-row method and under wilt 

 infection conditions, it was found possible to build up varieties 

 12 



