EXPERIMENT STATION BULLETINS. 561 



water a second cooling bath was employed. This bath was exactl}^ the 

 same in form as the above, except that its temperature was maintained 

 at about — 5 or — 6°C. It was found necessary to have this second cool- 

 ing bath with such high temperature below zero degree because water 

 would not freeze very readily at the temperature of — 2 or — 3°C of tbe 

 first bath, but it would at the temperature of — 5 or — 6°C. Also, since 

 the thermometer was calibrated with distilled water very often in the 

 course of a series of freezing-point lowering determinations it was found 

 very convenient and time-saving to have a separate cooling bath with the 

 proper temperature. 



The original procedure of determining the freezing-point lowering of 

 soils was considerably changed in the second investigation. In the 

 present investigation the following procedure was employed: First, the 

 thermometer was set so that the mercury thread would come to rest 

 toward the upper part of the scale when the bulb was held in a mixture 

 of pure ice and distilled water. The freezing point of distilled water 

 was then determined and this point was used as a standard for ascertain- 

 ing the freezing point lowering of soils. Then the freezing point of soils 

 was ascertained by placing in the tube B a column of soil of about an 

 inch in height and inserting the bulb of the thermometer into this column 

 of soil until it was completely covered. The tube containing the soil 

 and the thermometer was placed directly in the cooling mixture through 

 one of the holes in the cover and allowed to supercool to about 1°C. 

 Then by holding the tube near its mouth with one hand and with the 

 other moving the thermometer, solidification was started in the soil. 

 Immediately upon the commencement of solidification, as indicated by 

 the rise in temperature, the tube was at once taken out of the cooling 

 mixture and put in the air jacket. The temperature was allowed to 

 rise until it came to rest at a certain point on the scale and remained 

 at that point for some time. This maximum temperature obtained was 

 considered as the proper freezing point of the soil and recorded. The 

 thermometer was always tapped before taking the finaj reading. 



By placing the soil to cool directly in the cooling mixture instead of 

 in the air jacket as was done previously, the desired supercooling was at- 

 tained in a very short time, usually in about 10 minutes. Also by main- 

 taining the temperature of the ice mixture at — 1 or — 3°C the loss of 

 heat from the soil was considerably less than that at — 5°C as used in 

 the first investigation, and consequently the error, if any, in the results 

 was smaller in the low temperature below zero than in the high. 



If the temperature of the bath was greater than — 2 or — 3°C, solidifi- 

 cation would start in the soil before supercooling had taken place, and 

 of course the proper freezing point of the soil could not be accurately 

 ascertained. But if the temperature of the cooling mixture was not 

 greater than the above limits no difficulty would be experienced in super- 

 cooling the soil. 



It was not found advisable to reduce the decree of supercooling. Al- 

 though 1°C supercooling is considerable, it is believed that no serious 

 errors arise therefrom. 



For preparing samples of soil to determine their freezing-point lower- 

 ing the following method was employed : For low moisture content about 

 20 grams of air dry soil was taken in the case of sandy loams, silts, and 

 clays, about 30 grams of sands, and about 10 grams of mucks and peats, 

 and mixed well with the desired amount of water. The mixing consisted 

 71 



