Dec. 15. 1914 Oil Content of Seeds and Nutrition of Plant 229 



PROBLEMS INVOLVED AND GENERAL METHODS OF PROCEDURE 



The growth, development, and composition of the individual plant 

 depend on the combined action of heredity and enviromnent. It fol- 

 lows that in any investigation of heredity or of environment in relation 

 to plant life such methods should be followed as will make it possible 

 to distinguish between these two forces. This principle of clearly dis- 

 tinguishing between the effects of environment and those due to heredity 

 is simple enough in theory, but in practice it is often extremely difficult 

 to follow because of interrelations between the two forces which are not 

 fully understood. Thus, nouheritable variations in the progeny of 

 relatively pure strains grown under what would seem to be uniform 

 environmental conditions become especially apparent when we come to 

 deal with quantitative differences such as are involved in a study of 

 variable chemical composition as influenced by the factors of nutrition. 

 In our experiments on the nicotine content of tobacco, the oil content of 

 seeds of various species, and other quantitative characters, efforts to 

 insure the greatest unifonnity possible in environmental conditions have 

 failed to effect anything like uniform composition in the progeny of 

 individuals representing the purest strains available. Some of the 

 strains of tobacco under study had been inbred for eight generations. 



It follows that in dealing with quantitative differences in composi- 

 tion as influenced by environment a sufficiently large number of indi- 

 viduals must be used to avoid misleading results due to variations that 

 can not be brought under control. Where conditions have made it 

 impracticable to grow relatively large numbers of plants for study, 

 the alternative of repeating the experiment has been adopted. It is 

 equally important to use caution in reaching generalizations based on 

 results obtained with an insufficient number of species or varieties. 

 In the course of our work on the oil content of seeds it has been fre- 

 quently observed that in the case of the soy bean, for example, different 

 varieties arc not always influenced in the same manner by the environ- 

 ment. In such plants as the soy bean and cotton, therefore, as many 

 varieties as practicable have been included in the experiments. In 

 selecting plants for study those that are of special importance by reason 

 of the oil content of the seeds and which are otherwise adapted to the 

 work in view have been taken. Thus far we have utilized cotton (Gos- 

 sypium spp.), soy bean, peanut (Arachis hypogaea), and sunflower. 



With reference to chemical composition, the commercial value of seeds 

 evidently depends on the relative percentages of their several constituents, 

 but from the standpoint of the grower the returns also depend on the 

 quantity of seed produced — that is, on the number and size of the seeds 

 produced by the individual plant or on a given acreage. Hence, in the 



