THE VALENCY RULE AND THE HOFMEISTER SERIES 67 



per ounce; 27 per cent alcohol, U. S. P. IX) was used. Erlen- 

 meyer flasks of a volume of about 50 c.c. were rinsed with 95 per 

 cent alcohol and then filled to the neck with the collodion solution. 

 After the flask was filled with collodion, the latter was allowed to 

 pour out slowly from the flask which was rotated slowly by hand 

 during this process. The process of rotating the flask and pour- 

 ing out the collodion was timed to occupy exactly 2 minutes. 

 Then the Erlenmeyer flask, which was now empty except for a 

 film of collodion adhering to the inside of the glass wall, was 

 allowed to dry for exactly 2 minutes at room temperature. It 

 was then put under the faucet and tap water was allowed to run 

 in in a gentle stream for 5 minutes. The collodion film formed 

 inside the flask could be pulled out being an exact cast of the flask. 

 These collodion bags were closed, with the aid of rubber bands, 

 by a conical rubber stopper which was perforated to allow a glass 

 tube to be pushed through. The collodion bag was filled with the 

 solution of protein with the aid of a small funnel, all air bubbles 

 were removed and the glass tube was pushed into the bag to serve as 

 a manometer. The bag was then put into a beaker usually con- 

 taining 350 c.c. of water, having the same pH as the protein solu- 

 tion. The surface of the stopper was so adjusted that it lay in the 

 surface of the water in the beaker and the glass tube (or manometer) 

 was pushed a little deeper into the bag so that at the beginning the 

 level of the protein solution was about from 20 to 30 mm. above 

 that of the water in the outside beaker. 



The water diffused from the outside beaker into the protein 

 solution and the column of liquid in the manometer rose to a 

 maximum which was usually reached in about 6 hours or possibly 

 less. It must be taken into consideration that two changes in 

 pH will occur in these experiments which affect the osmotic 

 equilibrium. The one change is due to the Donnan equilibrium 

 which was referred to in the introduction. The other change is 

 due to the influence of the CO 2 of the air in the outside solution 

 and this influence is especially disturbing when alkaline 'solutions 

 are used. It must also be borne in mind that in these experi- 

 ments the protein solution is also usually diluted through the 

 entrance of water into the collodion bag. In later chapters 

 measures will be mentioned by which these sources of error 

 can be avoided or diminished. 



