322 



Journal of Agricultural Research 



Vol. XII, No. 6 



Table II. — Moisture equivalents, hygroscopic coefficients, specific heat, and mechanical 

 analysis of experimental soils — Continued 



MECHANICAIv ANALYSES 



[Averages of two analyses) 



In choosing the soils in the field the following plan was adopted. 

 One soil of each type was taken as the typical sample. Then each suc- 

 ceeding member of that group was chosen by comparison with a small 

 sample of this type. This same small sample was carried while all the 

 work was being done on that group, and constant comparisons were 

 made when there was ground for doubt. In this way it was believed 

 that each member of a group would approach very closely to the other 

 members in general physical texture. 



The samples in each case were taken from the top foot of soil. Pre- 

 liminary borings with a soil auger were made in order to ascertain whether 

 the subsoil was free from hardpan or other abnormal factors which might 

 influence the surface soil. The amount of each of these samples was 

 2>^ tons. The soil was shipped in new, heavy burlap sacks holding ap- 

 proximately 100 pounds each. Every attempt was made to prevent 

 undue drying or exposure to sunshine after the soil was sacked. 



Upon the arrival of the samples in Berkeley they were immediately 

 passed through X-iiich concrete sieves to obtain a uniform physical tex- 

 ture. This work was performed with all possible expedition, and it is 

 beUeved that they had a normal bacterial flora when placed in the 

 containers. None of the soils, except No. 14, was actually air-dry when 

 finally prepared for use. The ammonifying and nitrifying powers of the 

 soils have been determined as a test of normal biological activity. This 

 work was performed at the beginning and close of the season of 191 5 

 and at the end of the growing season in 191 6. The tests were carried out 



