12 W. A. CANNON 



full. With the cocks joining the reservoir and the container 

 open, so that water may flow between the two, the container 

 is raised or lowered sufficiently to bring the water in it to a depth 

 of about 2 cm. when the cocks are closed. A small plug of glass 

 wool is put in the bottom of the container, to prevent the escape 

 of the soil, and all air is removed from the wool by gentle pack- 

 ing. The apparatus is then ready for the introduction of soil. 



The actual manipulation, so far as is necessary to give it in 

 detail, is as follows: The container is filled with soil. The cocks 

 are open, permitting the rise of water and its flow out of the 

 reservoir, and the soil is at once run through several times with 

 a small glass rod. The last operation is for the purpose of pro- 

 moting the settling of the soil as well as to prevent the premature 

 collecting of air bubbles at the bottom of the soil container. 

 All of this time, it should be understood, the water is flowing 

 freely from the reservoir into the container and is rising in the 

 soil. As soon as the soil appears to be uniformly settled the 

 cock beneath the reservoir is so turned as to cut off the flow of 

 water from it and, at the same time, to connect the container 

 with the manometer. The position of the mercury is observed 

 at. this moment and the rise is watched until the upward move- 

 ment ceases, or until the mercury falls. The uppermost point 

 reached by the mercury is considered the measure of the capillary 

 pull of the soil under the conditions of the test. 



When the mercury falls the experiment is ended. If it is 

 desirable, as it generally is, to repeat the test, and if there are no 

 air bubbles in the bottom of the soil container, it can be done in 

 one of two ways. Either the soil can be removed and the test 

 begun anew, as must happen should air be found in the bottom 

 of the container,, or the following manipulation can be resorted 

 to. Water is immediately run into the container and the soil 

 is run through with a glass rod, after which the water is turned 

 off and the position of the mercury noted and its rise observed 

 as before. This operation can usually be repeated several 

 times before the collection of air bubbles in the bottom of the 

 soil compels the termination of the test. 



As illustrating tj^pical results the following experiments can 



