36 



digestion was continued for three hours after the solutions were 

 clear, so that the phosphotungstic acid would not interfere with 

 the accuracy of the determination. The "amino nitrogen in the 

 filtrate from the bases" was determined on duplicate portions of 

 10 cc. each by means of the Van Slyke (1912) apparatus. 



2. The method in detail for a mineral soil. Duplicate portions 

 of 250 grams were hydrolyzed in 500 cc. round bottom Kjeldahl 

 flasks for 48 hours on different sand baths. Allowance was always 

 made for the lime content of the soil, and the requisite amount of 

 hydrochloric acid added to insure the presence of a constant boiling 

 acid (sp. gr. 1.115), and a volume of approximately 250 cc. The 

 solutions boiled smoothly and gave no trouble by bumping. Air 

 dry soil was used in all cases, but the moisture was determined on 

 a separate portion and all data calculated to the dry basis. 



On completion of the hydrolysis each of the two samples was 

 diluted to a 1000 cc. in measuring flasks and allowed to settle for 

 at least 24 hours. The clear solution was then syphoned off and 

 an aliquot of 500 cc. analyzed according to the usual method of 

 Van Slyke. In nearly all cases this solution was straw color due 

 to the presence of ferric salts that had been formed during the 

 hydrolysis. No black color, the usual color of a protein hydrolysate, 

 was observed in any instance. 



Another aliquot of 100 cc. was used for the determination of 

 total nitrogen in the solution by making duplicate Kjeldahl de- 

 terminations on 50 cc. portions. A second aliquot of 100 cc. was 

 used for the determination of u jodidi numbers" (cf. p. 25) when 

 they were determined. 



The soil remaining in the measuring flask was washed free 

 from soluble nitrogen with a 1 per cent solution of potassium sul- 

 fate by decantation from tall soil beakers, the solution after set- 

 tling being syphoned off not oftener than twice a day. This meth- 

 od was employed in order to prevent the clay from forming a 

 colloidal suspension. Distilled water alone would remove all elec- 

 trolytes and allow a portion of the clay to remain in the solution 

 in colloidal suspension. It is known that suspensions of finely di- 

 vided clay carry a negative charge in pure water. Since it is 

 necessary for the complete precipitation of colloids to. have some 

 electrolyte present it was decided to use a 1 per cent solution of 

 potassium sulfate. The negatively charged colloid was thus pre- 

 cipitated by the positive ions of the potassium sulfate solution 

 and at the same time the salt would not interfere with the subse- 

 quent Kjeldahl determination. 



A concrete example of the thoroughness of this washing may 

 well be given. It will be noted that 700 cc. of the original hydrol- 

 ysate was syphoned off for the different analyses. This left a total 

 volume of 300 cc. of residue and solution to be washed by decanta- 

 tion with 1 per cent potassium sulfate. By the methods of calcu- 

 lation given in the following paragraphs it was found that the re- 

 maining- solution contained 0.1089 gram of nitrogen. If three- 

 fourths of the wash solution is removed each time, there will remain 



