378 University of California Publications in Agricultural Sciences [Vol. 3 



advisability of judging all results in the light of the probable error 

 is admitted, no attempt has been made to apply this factor to the 

 results reported in this paper. As the result of the effect which such 

 a factor might have upon the results of bacteriological determina- 

 tions carried on only in duplicate, or upon the results of greenhouse 

 work done in triplicate, one hesitates to draw conclusions, especially 

 those based upon minor variations. Hence in this work only the more 

 marked results will be considered of significance. 



When planning the work it was thought that three or four samples 

 of a type would be enough to show whether or not a given type was 

 approximately uniform, or widely variable, and as to whether the 

 types were similar to one another, or quite dissimilar. But it now 

 seems, after comparing the determinations run on the larger number 

 of samples of the Hanford and San Joaquin types, 9 and 8 respec- 

 tively, with the determinations run on the Altamont and Diablo types, 

 of which there were a much smaller number of samples, 3 and 4 

 respectively, that the larger series gives a much better insight into the 

 variations of a given type and affords a much better basis for con- 

 clusions. 



Hence, as regards the laboratory work thus far carried out, the 

 emphasis has been placed upon the Hanford fine sandy loam and the 

 San Joaquin sandy loam. Determinations have not been completed 

 on the Altamont and Diablo series to the extent that they have on the 

 former two. 



It is of no little significance that the Hanford fine sandy loam and 

 the San Joaquin sandy loam are very widely contrasted soils agricul- 

 turally. The Hanford is typical of good recent alluvial soil in this 

 state ; while the San Joaquin is typical of wide expanses of ' ' old valley 

 filling" soils that are considered poor as regards crop producing 

 power and are underlain by compact iron-cemented hardpan. Conse- 

 quently, the results of comparing soils so different from an agricul- 

 tural point of view, and so radically different as regards soil survey 

 criteria (though the textures are quite similar) will be of considerable 

 interest. They are of greater interest than the comparisons between 

 the Diablo and Altamont soils, as the latter are quite similar in agri- 

 cultural value and use, as well as in field appearances. Between the 

 Diablo or Altamont and the Hanford or San Joaquin one cannot 

 judge as closely regarding variations, for the soils are so radically 

 different. On the other hand, one can compare the soils of the heavy 

 and light types to see to what extent the chemical and bacteriological 

 results differ as compared with the physical results. 



