166 BREEDING CROP PLANTS 



variability. A careful survey of experimental studies does not 

 support this contention. The development of the shade- 

 grown tobacco industry in the Connecticut Valley is of interest 

 in this discussion. This shade method first originated in Florida 

 in 1896 and was tried experimentally in Connecticut through 

 cooperation of the Connecticut Experiment Station and officials 

 of the Bureau of Soils. In 1900 one third of an acre was grown 

 and the crop sold at an average price of 72 cents per pound. 

 A considerable acreage was grown in 1901 and the crop sold 

 at public auction at a much higher price per pound. Indis- 

 criminate introduction of unselected seed from Florida was 

 practiced and in 1902 over 700 acres were grown under shade in 

 Connecticut. The result was a disastrous failure, owing to a 

 lack of knowledge of methods of handling and to the use of un- 

 selected seed. By further study of handling and through careful 

 selection in which artificially self -pollinated seed was saved, 

 the industry was placed on a firm foundation. This latter work 

 was carried on by the Bureau of Plant Industry (Stewart, 1908). 

 A knowledge of Cuban methods shows that imported Cuban seed 

 is a mixture of many types. Some experiments have shown that 

 the breaking up alluded to is an expression of the different 

 hereditary qualities of the parental seed plants. In 1912 Hassel- 

 bring grew a number of pure lines of tobacco in Michigan which 

 he had formerly grown in Cuba. No evidence of breaking up of 

 type was observed and whatever changes occurred in a pure line, 

 owing to the new conditions, were uniformly exhibited in all 

 plants of the pure line. Similar conclusions were reached from 

 the immediate introduction of individual seed capsules of different 

 tobacco plants from Cuba and their subsequent growth under 

 shade in Connecticut (Hayes, 1914). Careful studies at Pusa, 

 India, convinced the Howards (1910a) that new conditions 

 did not cause a breaking up of type. They ascribed the apparent 

 variability of new introductions to cross-fertilization, which 

 was shown to occur frequently in tobacco. 



Although there have been some differences of opinion as to 

 the cause of variability of new introductions, there is uniformity 

 of belief regarding the methods of obtaining purity of type. 

 Artificial self-pollination gives uniformity, and continued 

 self-fertilization produces no harmful effects. This method was 

 strongly recommended by officials of the United States Depart- 

 ment of Agriculture (Shamel and Cobey , 1907) and by the different 



