372 GIBBS AND AGCAOILI. 



In the hands of the writers this method has not only failed as a 

 quantitative method for the estimation of coconut oil, but it has also 

 failed to show any marked differences, which can be depended upon, 

 between a number of different fats. The reason is easily found. 



The permanganic acid which is formed upon acidification of a potas- 

 sium permanganate solution is readily decomposed on exposure to light 

 or on gentle heating, with the separation of oxides of manganese and 

 loss of oxygen. On boiling the evolution of oxygen is more rapid.^ 

 Even a weak solution of permanganic acid continually evolves oxygen. 

 Dammer^ states that in the presence of an excess of sulphuric acid 

 permanganic acid is reduced. 



Morse, Hopkins, and Walker " have found that permanganic acid and potassium 

 permanganate are reduced by precipitated superoxide of manganese with the 

 liberation of three-fifths of the active oxygen and that solutions of potassium 

 permanganate are more stable if freed from suspended oxide and kept in darkness 

 or diffused light. Even pure solutions are decomposed in direct sunlight. Morse 

 and Reese " state that they have "always found dilute, moderately acidified solu- 

 tions of permanganate quite stable at ordinary temperatures, provided they were 

 free from oxide," and that the decomposition of permanganic acid by the peroxide, 

 attended by the liberation of oxygen, is a continuous reaction, which ceases only 

 when all of the acid has been reduced to the oxide. 



These references seem to have escaped the attention of Mr. Hodgson. 

 He mentions no precautions which were taken to purify his permanganate 

 solutions, does not speak of any decomposition of the permanganate and 

 altogether has no difficulty in obtaining results, which in view of our 

 knowledge of the behavior of permanganate solutions, are without suffi- 

 cient experimental foundation. 



Ross and Race " have found Hodgson's method to be "unworkable." Their 

 experiments have sho-ivn them that "sulphuric acid of the strength prescribed 

 exerts under the conditions laid do^^^l a considerable action on potassiimi perman- 

 ganate" and that "owing to the retention of the hydrated oxides of manganese by 

 the insoluble fatty acids liberated on the addition of acid" difficulty was experienced 

 in obtaining a good end point. Thompson and Tankard '- have found that the 

 permanganate solution is attacked by the reagents used and pronounce the process 

 "fundamentally unscientific and based upon error." 



When the method of oxidation of the saponified fats is carried out 

 according to the described method, the loss of active oxygen of the 

 permanganate solution varies little in the case of each of the fats and 

 oils with which we have experimented and moreover this loss in active 

 oxygen is about the same as when distilled water is used instead of the 

 soap solutions. In one case the lost oxygen escapes into the atmosphere, 



'Roseoe and Schorlemmer: Treatise on Chemistry (1900), 2, 919. 



^Handbuch der anorganisehen Chemie (1893), 3, 251. 



'Am. Chem. Jour. (1896), 18, 401. 



"Ato. Chem. Jour. (1898), 20, 526. 



^'■Chem. Kews (1908), 97, 110. 



'"Chem. Neics (190S), 97, 146. 



