3 *eb. s, 1917 
Measurement of Inactive Moisture in Soil 
199 
After the soil was placed in the bulb, the latter was completely filled 
with the ligroin. For expelling the air the soil was stirred with a rod 
until bubbles ceased to appear at the top. The process of stirring 
proved quite efficient in facilitating the removal of air. 
After all the air was expelled from the soil, the thermometer was 
inserted into the mouth of the dilatometer and a sufficient amount of 
ligroin was placed in the stem, so that when the mass in the bulb had 
cooled, the column of the ligroin would be at the lower part of the 
graduated portion of the stem. The stem was then covered by a special 
cape in order to prevent the loss of the ligroin through volatilization. 
The bulb of the dilatometer was then placed in a cooling mixture and 
allowed to cool. The cooling mixture consisted of crushed ice and 
common salt and was contained in a bath composed of two earthenware 
jars placed one inside the other and well insulated. The temperature of 
the cooling mixture was maintained at — 4 0 C. The dilatometer was 
allowed to remain undisturbed until the contents had attained the 
temperature of — 3 0 , that is, supercooled to 3 0 ; and then by taking 
hold of the dilatometer by the stem it was gently moved in the cooling 
mixture until solidification began. The beginning of the solidification 
was indicated by the rise of the ligroin in the stem. The bulb of the 
dilatometer was allowed to remain in the cooling mixture, with a frequent 
movement until the rise of the liquid in the stem ceased. The time 
required for equilibrium to be attained varied with the type of soil, but 
it generally took about half an hour. This length of time, however, 
could be considerably reduced by moving the bulb in the ice mixture 
very often. The total rise of the ligroin in the stem was taken to repre¬ 
sent the total amount of expansion due to the formation of ice. 
Whenever it was desired to study the effect of successive freezings 
upon the amount of water that freezes, the bulb of the dilatometer was 
taken out of the ice mixture, wiped dry, and held between the hands 
until the soil was thawed, as indicated by the rise of the temperature 
on the thermometer; then it was placed back into the cooling mixture 
and the process described above was repeated. 
For obtaining concordant results it was found essential to supercool 
the soil to the same degree. As will be shown subsequently, the amount 
of water that will freeze in a soil increases with the increase in super¬ 
cooling. This is especially true of colloidal soils. 
It was not found practicable to maintain in the bath a lower tem¬ 
perature below zero than — 4 0 C. * If the temperature was lower than 
— 4 0 , solidification would start before supercooling had taken place and 
no satisfactory results could be obtained on expansion. Furthermore, 
if the bath was too cold, the tendency for contraction would be too 
great, and the total expansion might not be indicated in the stem. 
