120 EXPERIMENT STATION RECORD. 



tlu' i)latiiiimi li;i(l settled the siipernateut liquid was decanted on to a 

 filter, the platiinini washed with nitric acid (1 part of acid to 9 of 

 water), and finally with boiliny water, dried, and weighed. 



At the beginning of the investigations the nitrogen was determined 

 iu 10 gni. of the soil by combustion with soda-lime. Tiie Kjeldahl 

 method was adopted afterward as more rapid and equally accurate, 

 the same weight of soil being used as in the other method. 



Tlie methods used in some cases for the determination of magnesia 

 by means of uranium solution and of iron by means of i»otassium per- 

 manganate are described in detail. 



In deternuning- ammoniacal nitrogen 20 gm. of the air-dried soil was 

 extracted with 100 <;c. of distilled water, 50 cc. of the li([uid distilled 

 with magnesia, and the ammonia determined in the distillate by means 

 of Nessler solution. 



Nitric nitrogen was determined in a similar extract by means of stand- 

 ard diphenylamin and nitrate solutions, and nitrous nitrogen by means 

 of Griess's reagent prepared by dissolving at the boiling point 1 gm. 

 of sulphanilic acid and 1 gm. of naphthylamin in 200 cc. of water 

 containing 2."> cc. of hydrochloric acid. The coloiation observed was 

 comi)ared with that obtained by the use of a standard solution of nitrate 

 ofi)otash. 



\'arious quick methods for determining the amounts of i)otash, phos- 

 phoric acid, and lime soluble in water or weak regeants are also noted. 



A method is described of estimating the amount of fertilizing ingre- 

 dients per hectare of soil, taking account of the inert gravel and coarse 

 soil which are not included in the chemical analysis. 



Moisture and absorptive power of soils. — Investigations relating to 

 the moisture of the soil were carried out in vegetation pots at two dif- 

 ferent periods in the course of the above soil investigations. The more 

 recent of these investigations show that from April 14 to July 18 the 

 subsoil of a bare soil stored up 12,658 liters of water per acre and that 

 it lost 5,813 liters, showing an excess of 0,845 liters, which is greater 

 than the average monthly rainfall during the period. The evaporation 

 from the bare soil from May 4 to July 18 was 1 9,757 liters, from a covered 

 soil 27,441 liters; but at the same time there was stored up in the sub- 

 soil of the covered soil 7,099 liters, as against only 2,716 liters in the 

 subsoil of the bare soil. The subsoil api)ears to be a reservoir which 

 stores up the excess of moisture in the soil in periods of heavy rainfall 

 and gives it out when needed in periods of drought. 



The nature of the substances removed from soils by leaching with 

 city service water and with solutions containing nitrate of soda and 

 superphosphate was studied in the following manner: An iron cylin 

 der 1.2 m. long and 2 dcm. in diameter, terminating iu a cone below 

 fitted with a lubber tube and stopcock, was partly filled with washed 

 gravel on which was placed 20 kg. of soil extending to a height of about 

 1 m. The liquids were poured on to the surfice and after traversing 



