^^2 Journal of Agricultural Research voi. xii. No. 6 



Then cool the candles and pass distilled water through them to remove charred 

 organic matter and salts formed by the process of ignition. 



In the work carried out in 191 5, with the analytical methods of the 

 Bureau of Soils, 150 gm. of fresh soil constituted the sample treated 

 with 750 c. c. of distilled water. In 191 6 the modified procedure which 

 has been described required a larger volume of solution. Accordingly 

 340 gm. of soil has been treated with i ,700 c. c. of distilled water. This 

 volume of solution is the maximum quantity that can be filtered in two 

 cylinders of the filtering apparatus. These two portions of solution 

 are afterwards united and furnish a volume sufficient for the determina- 

 tion of phosphates, calcium, potash, nitrates, magnesium, and total 

 solids, if desired. A moisture determination was in each case performed 

 on the sample of fresh soil, and the result used in correcting the analytical 

 data to bring it to a uniform basis. 



In this connection it should be pointed out that the form in which 

 the final results are to be stated is a matter of some importance. Mitscher- 

 lich (42), King {35), Whitney and Cameron (<5i), all expressed their 

 results on the basis of percentage or parts per million of the dry soil. 

 Gola {18), on the other hand, referred his to the concentration of the 

 solution obtained, which he considered to be representative of the soil 

 solution. Inasmuch as it is the soil solution which is our final object, 

 it would be natural to refer the analytical results to the actual soil 

 moisture determined in each sample. Desirable as it is, it is not believed 

 that we possess the necessary information in regard to the state of the 

 soil moisture to do this correctly. In discussing the results on the 

 moisture equivalents given in Table II we have referred to the unfree 

 water present in soils. 



The work done by Briggs and Shantz (<?) on the wilting coefficient of 

 soils has shown the variations which occur among different soils, espec- 

 ially where they are of diverse types. Their work demonstrates clearly 

 that a large amount of moisture is unavailable to the plant in soils 

 which are high in colloidal material. Recently Bouyoucos {4) has 

 called attention to the correlation which exists between the moisture 

 held in unavailable form in the soil, as determined by the above method, 

 and the unfree moisture determined by the dilatometer method. He 

 has also compared the results obtained by these two procedures with 

 the determination of the moisture which fails to freeze. A remarkable 

 agreement is found to exist. Working with some 14 soils used in the 

 present study Hoagland (27) has found that in all the silty clay loams 

 a large percentage, varying from 13 per cent to 18 per cent, does not 

 freeze. In the fine sandy loams the percentage is much smaller, from 

 4 per cent for soil 8 to 8 per cent for soil 10. The extent of these varia- 

 tions and their significance is discussed at greater length by Hoagland 

 (27). If this moisture is held, either loosely chemically combined or 

 physically absorbed or both, it is at least possible that it is not available 



