1912. 



PUBLIC DOCUMENT — No. 31. 



119 



using these volume i)crceutagcs as weight percentages, which 

 may be done, as experiment has shown that in the small tube the 

 error is not great enough to be taken into consideration. 



Very fine sand = 2.34 X 42.85 per cent. = 1.00 + grams. 

 Silts = 2.34 X 21.43 per cent. = .49 + grams. 



Clay = 2.34 X 35.72 per cent. = .84 + grams. 



Thus we have the weights of the very fine sand, silts and clay, 

 and by folloAving the same system used in calculating the i)ev- 

 centages of the sands obtained by sifting in the whole sample 

 we get the percentages of these constituents. 



Below are given results of several typical soils which were 

 analyzed by the long method and by the short method. It will 

 be seen that the results vary but little and that for a close ap- 

 proximate analysis the resiilts are accurate enough to warrant 

 the use of this method where time is an important factor. 



A criticism of this method may be raised, but its accuracy 

 and ease of manipulation cannot be doubted, as it has been re- 

 peatedly proved to give as good results for general use as the 

 long method, and in about half the time. Soils were analyzed 

 by outside parties, and then the same soils Avere analyzed in the 

 laboratory by the shorter method, and the results were well 

 within the acknowledged limit of error, as can be seen from the 

 following table : — 



Table showing the Results obtained on Various Soils from Analyses by the 

 "Long Method" and by the New Method. 



[Per cent, of organic matter, gravel, sand, silt and clay in 20 grams of soil.] 



