- 41 - 



correct the "day-by-day" discharge rates but that the total dis- 

 charge over a period of a "week" was acceptable. Since the 

 points of discharge in Test B are so close together and united 

 in a cojnmon portion of the detailed circulation no significant 

 error in the equilibrium distribution of bleed water could occur 

 as a result of small proportioning errors. In Test C, however, 

 a small proportioning error would be highly significant in the 

 equilibrium distribution of bleed water because the dilution canal 

 is part of a circulation pattern that is distinct frcm the Humble 

 Canal system. Test C was, therefore, not run until a reliable • 

 rotary proportioning device had been built and installed. The 

 total discharge of the bleed water pump was monitored as before 

 by timing the revolutions of the driving motor shaft and screw. 

 The results obtained in all bleed water discharge tests are be- 

 lieved to be representative and consistent in spite of the diffi- 

 culties enumerated. 



Two other pumping systems had significance in the model tests 

 (1) the dilution system which withdrew scaled volumes of water 

 from Lake Grande Ecaille and discharged them as a mixture with 

 properly scaled volumes of bleed water added, and (2) the rainfall 

 pumping system. The bleed water dilution system consisted of a 

 smaU gear pump driven by a separate electric motor and monitored 

 by counting the number of drops discharged per minute as they fell 

 frcm a calibrated dropper tip. Since the scaled volumes involved 

 a drop every few seconds it was possible to observe the rate of 



