598 STATE BOARD OF AGRICULTURE. 



point of the soil and recorded. The thermometer was always tapped 

 before tnkinc; the final reading. 



It was discovered very soon at the beginning that it required a very 

 long time for the soil to cool in the air jacket, owing to the poor heat 

 conductivity of the air in the jacket. It was finally decided to hasten 

 the process by bringing the tube containing the soil and the ther- 

 mometer in intimate contact with the cooling mixture and allowing 

 the soil mass to cool to about 5°C above the freezing point of water, 

 or until there was only a small amount of mercury at the upper part 

 of the reservoir, and then taking out the soil tube from the cooling 

 mixture and inserting it at once into the air jacket and letting the 

 soil mass to cool slowly. This procedure, however, was generally success- 

 ful when the soils contained a high moisture content, but not when 

 they had a low water content. In the latter case solidification would 

 commence before supercooling had taken place and the proper lowering 

 of the freezing point could not, of course, be determined. The plan, 

 therefore, was modified, and in its final form consisted of cooling the 

 soil directly in the cooling mixture until all the mercury had left the 

 reservoir and then taking that portion of the tube containing the soil 

 between the palm of the hand and warming it until the mercury thread 

 began to rise again. The tube was then placed into the air jacket in 

 the cooling bath and allowed the soil to cool gradually. The process of 

 warming the soil slightly after it was taken from the cooling mixture 

 proved a great success as it almost always prevented the occurence of 

 solidification before supercooling had taken place. With the above plan in 

 its final form, therefore, the time required for making a freezing point 

 determination was reduced about 60%, and as many as 30 freezing 

 point determinations have been made in one day — 8 hours. 



For taking off the initial excess of heat from the soil system it was 

 found convenient to employ a second battery jar containing a cooling 

 mixture of the same temperature as the first. 



It was found very essential to allow the soil to supercool to about 1°C 

 below its freezing point, although solidification could be induced at 

 the supercooling of only about 0.3°C, in order to obtain quicker and 

 more accurate readings. By supercooling the soil to about 1°C, suffi- 

 cient solid or crystal ice of the pure soil water is immediately sepa- 

 rated to warm the soil system quickly up to its proper freezing point. 

 If the amount of the supercooling is small, the separation of the solvent 

 is very slow, the thermometer comes to rest only very slowly, and the 

 readings obtained are not accurate, especially with soils of low moisture 

 content. 



On the other hand, too much supercooling is also a source of error, 

 and especially if the soil solution is highly concentrated. The freez- 

 ing point of a solution is the temperature at which ice and solution exist 

 in equilibrium. Upon freezing, pure water separates in the solid phase, 

 and the liquid which remains, and whose temperature in equilibrium 

 with the mass of the ice crystals, is read from the thermometer, is not 

 the same, but more concentrated than the original solution. In con- 

 sequence, the observed depression of the freezing point is lower than 

 the true lowering of the freezing point of the solution under investi- 

 gation. The amount of solid solvent formed depends upon the amount 



