Dec. IS. 1914 Oil Content of Seeds and Nutrition of Plant 243 



In addition to these field experiments with the soy bean, a number of 

 tests on different soil types were made under specially controlled condi- 

 tions. In the spring of 191 2 the two varieties designated as S. P. I. Nos. 

 30599 and 30745 were grown on the above-mentioned heavy clay and 

 Potomac Flats soils placed in the greenhouse at the Arlington Experiment 

 Farm. The two soils occupied adjoining portions of the same bench, 

 received the same quantities of water, and were exposed to the same 

 conditions of light, etc. In 191 1 and 1912 soy beans were grown in differ- 

 ent soil types contained in large earthen pots set into the soil. These pots 

 consisted of glazed tiles 3 feet long and 18 inches in diameter. A per- 

 forated bottom of concrete was set in each tile cylinder 3 inches from the 

 lower end. The tiles were then set into the soil so that the upper ends 

 extended 3 inches above the ground level. The cylinders were filled to 

 the ground level with the different soils. Series of tiles were thus installed 

 in the clay soil of the Arlington Farm and in the Norfolk fine sandy loam 

 in the vicinity of Manning, S. C. At each location one-half the total num- 

 ber of cylinders were filled with the native soil and the second half with 

 soil transported from the other point. In preparing the soils for the tiles 

 the surface soil and the subsoil were collected separately and each lot 

 very thoroughly mixed. In each case a sufficient quantity of the native 

 soil thus prepared was shipped in bags to the other point, so that pre- 

 sumably at each point there was a series of insulated cores of two very 

 different soil types embedded in the native soil. Apparently the insula- 

 tion from the surrounding soil should be practically perfect, except pos- 

 sibly as regards temperature, and a special experiment, in which a cylinder 

 was inclosed in a second one of much larger size and filled with the same 

 soil, indicated that the temperature of the surrounding soil had but little 

 effect on that contained in the cylinder. The two soil types, of course, 

 would be exposed to exactly the same weather conditions in any par- 

 ticular locality. A second series of cylinders was installed at the Arling- 

 ton Farm and filled with the native clay soil and with Hartford fine sandy 

 loam from South Windsor, Conn. In 191 2 an additional set of cylinders, 

 filled with the Potomac Flats soil, was used in connection with the first 

 series. 



In the first series of cyHnders the variety designated as S. P. I. No. 

 17852B was used in 1911, while S. P. I. No. 32907 was used in 1912. In 

 the second series S. P. I. No. 21755 was used. The results of the various 

 tests with soy beans are summarized in Table XI. Each test is com- 

 plete in itself and of course can not be compared directly with the others, 

 since, with one exception, different varieties were used and the climatic 

 conditions were not the same. The results of each separate test are 

 based on data from a considerable number of individuals of each variety, 

 usually 50 to 100, and in most cases are the averages from several varie- 



