FLAX AND TOBACCO 167 



state experiment stations. The Howards, in India, likewise 

 urged the use of self -fertilized seed. Garner (1912) states that 

 several types have been inbred by growing the seed under bag 

 from six to eight years without any observable change in vigor 

 or habits of growth. These facts, together with the studies 

 of inheritance of quantitative characters, show that the pure- 

 line theory and Mendel's law furnish a reliable guide to tobacco 

 breeding operations. 



As quality of cured leaf is of such great importance in tobacco, 

 it is necessary that the breeder have a thorough knowledge of the 

 sort of leaf desired. Practical breeding operations must then be 

 carried on under the soil and climatic conditions in which the crop 

 is to be grown. An added complication is the necessity of basing 

 the final judgment of a particular selection upon the comparative 

 value of the cured leaf after fermentation. The difficulties of 

 comparing numerous strains, while not insurmountable, are 

 naturally much greater than for an equal number of small grain 

 selections. 



Mutations in Tobacco. The sudden appearance of giant 

 plants with abnormally high leaf number has been recorded in 

 the Sumatra, Maryland, Cuban, and Connecticut Havana varie- 

 ties of N. tabacum (Allard, 1916). These new forms under field 

 conditions have a much longer period of vegetative vigor than 

 the normal varieties. Consequently blossoming does not take 

 place under ordinary field conditions. Otherwise the general 

 habit of each of these new types is not very different from the 

 normal variety from which it was obtained. 



Two of these new varieties of giant habit are of some commer- 

 cial importance. A short account of their first recorded appear- 

 ance together with their cultivation as commercial varieties will 

 be given. Giant plants were noted in 1912 in the Cuban variety 

 which is grown under shade in the Connecticut Valley. (Hayes 

 and Beinhart, 1914). The history of the normal Cuban variety 

 from which the giant type was obtained is of interest (Hayes, 

 1915). Seed of the normal variety was saved under bag, which 

 insures self-fertilization, from 1904 to 1909 inclusive. In 1910 

 and 1911 seed was saved in bulk from plants which were grown 

 under the cheese-cloth cover used in producing shade-grown 

 tobacco, but individual plants were not bagged. During the 

 period from 1904 to 1910 no abnormal types were observed. 

 Studies of leaf inheritance in the Cuban variety were made from 



