32 



IOWA ACADEMY OF SCIENCE 



same temperature as the solution. The results have been so satis- 

 factory that a statement describing the method apart from the 

 main subject of investigation seems to be justified. 



The arrangement of the whole apparatus is shown in figure 1. 

 About ten gallons of water are contained in a tank of sheet iron 

 covered with layers of asbestos, felt and canvas. The regulator is 

 the ordinary form of Ostwald, designed for heating with gas. 

 The vessel for the solution is a large test-tube and it contains 

 a pipet holding about 20cc. as shown in figure 2, A. A large 

 rubber stopper was cut through to one of the holes, the shank 

 was inserted into the hole and the cut was closed with rubber 

 cement. A second pipet for higher temperatures where the solu- 



i 



f— J5C«— •] 



B 



Fig. 2. 



tion became more concentrated, is shown in figure 2, B. The 

 meter shown in figure 1 is a test gas meter reading to .001 cubic 

 foot. 



At first the thought was to stir both the bath and the solvent 

 with the same stream of air, but this introduced undesirable 

 complications and two streams were used. Ordinary air from 

 the reservoir was passed into the bath at the bottom and near 

 the middle at the rate of about three cubic feet per hour, which 

 was found ample as shown by two standard thermometers whose 

 bulbs were placed in different parts of the bath for comparison. 



