87 



Since the tube M has its lower end immersed in about two inches 

 of fluid, a definite pressure is required to force air from it. The 

 lower end of tube J is sealed by the water in K. Therefore, when 

 a drop of water from E falls into the tube F it forms a slug which 

 slightly compresses the air trapped in the system and the slug 

 comes to rest a little below the throat of the thistle. The next drop, 

 falling squarely into the upper end of F traps a slug of air, and 

 the first slug of water moves down further compressing air in the 

 system, while the second slug slides into the tube above the inter- 

 cepted air. The fourth or fifth drop will so increase the pressure 

 that the air in tube L-M will burst through the liquid in the flask 

 N in a series of bubbles, agitating the contents of the flask and 

 allowing the slugs of air and water in F to travel downwards rapidly 

 until the reduced pressure is again balanced by the fluid over the 

 bottom end of M. Slugs of water which pass into the "tee" G flow 

 downwards into J and overflow from K into T, 



Some simple adjustments may be necessary. A nice balance 

 between the pressures in the two lower tubes J and M is required. 

 The immersed length of tube J must exceed the immersed length 

 of M by an inch or so, to ensure sufficient back pressure to overcome 

 the resistance in M and permit some oscillation in the air volume. 

 The flask K, therefore, should be not less than 6" tall. The rela- 

 tive vertical positions of K and N may require adjustment by set- 

 ting one or the other on a wooden block. The greatest difficulty is 

 to ensure that no water from F passes into L and thence into N. 

 Any such addition to N will spoil the adjustment of the nutrient 

 fluid. If the "tee" G is left in a vertical position the capillary action 

 of the branch and the velocity of air in the tube will carry a part 

 of every water slug into L. By moving the whole of the bottom 

 set-up sideways and securing the tube M at a correct height by 

 means of the spring clothes-pin P, it was found possible to cause 

 the water slugs to break in G so that the liquid ran down the side 

 away from the branch. The film on the branch side also broke at 

 the entrance to the branch and any water splashed into the branch 

 flowed back into J. Any contamination of the water causing soapi- 

 ness will also cause water to be carried into the branch through 

 the film not breaking at the entrance. 



It is advisable not to cut the grooves in the plug E too deeply 

 at first ; gradually deepen or add grooves as found necessary in 



