116 EXPERIMENT STATION. [Jan. 



call it, the " long " method, so it became necessary to devise a 

 method which would materially shorten the process and still give 

 accurate results within a reasonable limit of error. 



After considerable experimentation a satisfactory method was 

 devised and has been used with success in our work here the past 

 year. It is not claimed that this method is absolutely accurate, 

 nor is any for that matter, as the limit of error, even when using 

 the most approved centrifugal methods where the greatest care 

 is used, is admittedly large, dependent somewhat of course on 

 the manipulator. 



A great number of comparisons have been made of the results 

 obtained by analysis of soils by the long method and the short 

 method and are given in the following pages. The method used 

 by us is more or less of an adaptation of the centrifugal method 

 in general use. 



In brief, the centrifugal method in general use is as follows: 

 the soil is carefully sampled and a part of the sample which 

 passes through a 2-millimeter sieve is used for analysis. Five 

 grams are usually taken and dried at 110° C. This sample is 

 then shaken with water, to which a few drops of ammonia have 

 been added, for six hours or more. The sample is then placed in 

 tubes and centrifuged until all but the clay particles have sub- 

 sided ; these, with the water, are then decanted off and evapo- 

 rated to dryness and weighed. The silts are found by allowing 

 everything larger in size than .05 millimeter to subside, de- 

 canting the liquid, evaporating, drying and weighing. The re- 

 maining sands are dried and weighed and then sifted bj' four 

 sieves into five grades. The organic matter is determined 

 usually by the chromic acid method, but should not be con- 

 founded with the " loss on ignition " which is often erroneously 

 termed organic matter. 



This process, as can plainly be seen, takes a long time to carry 

 through, and is not applicable where quick results are desired. 



The briefer method in use at this station is as follows : the 

 sample of soil as brought to the laboratory is first thoroughly 

 mixed and then dried at 110° C. It is then sifted through a 

 2-millimeter sieve and all that passes through is classed as soil. 

 This is again mixed and 10 grams taken for analysis. This is 

 heated to obtain the " loss on ignition," in a platinum or porce- 



