EXPERIMENT STATION BULLETINS. 515 



From Table III it will be noted that the amount of moisture ex- 

 tracted depends upon the kind of soil. With some soils the percentage ex- 

 tracted is very high. For instance, from the sandy soils Nos. 32 S, 33 

 and 34 of B, from 59 to 74 per cent of the moisture content of the soil, 

 No. 35 of A (7.6G%) 35 percent of the soil moisture was removed and 

 considering the low moisture content, it is to be seen that the extraction 

 was very efficient. 



From the column showing the number of c.c. extracted, it is evident 

 that sufficient amount of solution is obtained in nearly every instance 

 from a single cylinder for analytical purposes. In the majority of the 

 extractions the larger cylinder gave better results, — more solution in 

 proportion to the amount of soil usetl than the snmller one. In some 

 cases there was not much difference in the results. 



In a series of extractions with Soil 18, a variation from the usual 

 practice occurred, which showed further possibilities of the method. 

 Soil No. IS of A was received from Saginaw in a nearly air dry condi- 

 tion with a moisture content of G.25 percent (oven dried basis). Suf- 

 ficient ammonia-free distilled water was added to this soil the night be- 

 fore the extraction was made to make a total moisture content of 

 29.74%. After the extraction was made, that portion of the soil un- 

 touched with oil was removed and served for a second extraction (18a) 

 on the same day. The amount derived as a result of both extractions 

 equals 43.71 percent of the total moisture content present at the begin- 

 ning of the first extraction. The percentage of moisture in 18a corres- 

 ponds very closely to the average of the moisture equivalent obtained for 

 the small and large cylinders of Soil 18. From Soil 3 of C, a SaginaAV 

 soil, whose moisture content was 11.56 percent no solution could be ob- 

 tained. Ammonia-free distilled water was added and extraction was 

 made and results were obtained as noted in the table. Soil 3a was an- 

 other portion of the same soil v.-hose moisture content was raised 8 

 months later and extracted. Soils 4 and 4a of I) Avere the same soil ob- 

 tained at the same time, stored in large galvanized iron pails paraffined 

 to keep tlie soil from the sides of the pail and Avas covered with paraft"n 

 to prevent loss of moisture. The extractions Avere made at different 

 times. Soils 39 of F and 40 of G were obtained a week before the ex- 

 tractions Avere to be made for the purpose of dividing each sample into 

 two portions and raising the moisture content of one portion of each 

 sample. LikcAvise soils 42 of F and 43 of H Avere obtained, but no change 

 AA'as made in its mosture content. Soil 41 of G Avas obtained two days 

 before the extraction, Avhile all the others Avere used on the day of col- 

 lection. 



The moisture equivalent obtained, especially of those sandy soils with 

 a low percentage of organic matter present, approached quite closely 

 that calculated. This is notable in 35 of A, 32, 33, and 34 of B. Most of 

 them are Avithin the limit of error 4:^ 3 alloAved by Briggs and McLain\ 

 This is not true of the other soils. Perhaps this may be partially ex- 

 plained by the fact that the oil penetrated the former soils, but did not 

 penetrate the latter soils. The moisture equivalent of the peat and muck 

 soils is very high. From such high moisture content one might expect 

 a greater volume of the solution than was obtained, but, it indicates how 

 tenaciously the soil Avater is held by the organic matter. Another fact 



^Briggs, L. J. and McLanc, J. AV. Ibid. p. 18. 



