32 DIVISION OF WATER RESOURCES 



allowing' water to fiow to the soil tank. As the water level drops in 

 the supply tank a partial vacuum is formed above the water surface 

 and the water drops in the vent tube from the original level to a point 

 depending- upon the degree of vacuum established. This point is 

 determined by the ditference in the pressure heads, due to atmospheric 

 pressure and the partial vacuum in the supply tank. Water will con- 

 tinue to fall in the vent tube, but at a greater rate than in the Mariotte 

 tank, until the pressure head corres])onding to the atmospheric pressure 

 minus the pressure head caused by the partial vacuum is balanced 

 by a column of water equal to the ditference in elevation between the 

 water surface in the Mariotte tank and the bottom of the vent. Water 

 will then stand in the vent at the bottom of the tube with the pressure 

 at this point atmospheric. 



If the water continues to flow, air will enter the glass gage through 

 the vent tube, bubbling upward through the water in the gage to enter 

 the top of the supply tank. Water will continue to rise in the soil tank 

 up to the level of the lower end of the vent, at which point the atmos- 

 pheric pressure in the soil tank and the bottom of the vent tube is the 

 same. As there is no ditference in pressure and both points are at the 

 same level, there is no head to cause further flow and bubbling will cease. 

 AVhen the water table in the soil falls below the bottom of the vent, 

 the balance of pressures is again disturbed and a flow of water will 

 again start from the Mariotte tank, replacing the amount used. 



Some of the difficulties in the accurate use of the Mariotte tank 

 should not be overlooked. As the partial vacuum in the tank must be 

 maintained at all times, pipe connections must be air-tight. Air leaks 

 through the many joints in the system disturb the balance of pressure 

 necessary for automatic control. Thorough insulation against tempera- 

 ture changes inside the tank have been previously mentioned. Such 

 changes cause expansion or contraction of the tank itself, of the water 

 in the tank and also of the air in the chamber above the water. The 

 combined result is to cause changes in the vapor pressure with a result- 

 ing influence upon effective regulation. 



Water in the glass tube will fall with an increase in temperature 

 in the Mariotte tank, and readings on the scale taken at this time will 

 be erroneous. A test of the effect of temperature on scale readings 

 showed that an increase of 30° F. in air temperature caused a fall 

 of 1 cm. in the water surface in the glass gage. The temperature change 

 inside the Mariotte tank during the test was not measured. When the 

 temperature returned to the starting point the water in the gage 

 returned to its original position. If the gage readings are taken at an 

 early morning hour each day, the difference in the readings will repre- 

 sent a true rate of loss as early morning temperature changes are too 

 small to affect vapor pressure. Eeadings taken at other times of the 

 day may be in error unless complete insulation of the supply tank is 



effected. 



SAN BERNARDINO STATION 

 Description of Site 



In choosing a site for tank experiments in the upper Santa Ana 

 River Valley, a small plot on the grounds of the Antil Pumping Plant 

 of the San Bernardino Water Department,* about 1 mile east of town, 



* The Municipal Water Department of the City of San Bernardino, tlirougli tlie 

 courtesy of William Starke, as.sisted in maintaining this station. 



