EFFECT OF AI.KAI.I TREATMENT ON" COCOAS. 7 



ash of a cocoa treated with 2 per cent of a mixture of magnesium 

 and potassium carbonates. The maximum ash, 11.62 per cent, is 

 that of a cocoa treated with 1.5 per cent potassium carbonate and 

 1.5 per cent sodium carbonate. As would be expected, the ash 

 varies approximately with the amount of alkali added. 



Water-soluble ash varies from 2.43 to 9.36 per cent, with an aver- 

 age of 6.35 per cent. The lowest soluble ash is that of a cocoa which 

 is treated with magnesium carbonate, which, of course, is insoluble 

 in water. Next to this is one with a soluble ash of 2.87 per cent, 

 which has been treated with ammonium carbonate. There are sev- 

 eral varying from 3.52 to 4.75 per cent, which have been treated with 

 a mixture of magnesium and potassium carbonates. The lowest 

 water-soluble ash of those which are treated with soluble nonvolatile 

 alkalies is 4.89 per cent. 



The water-insoluble ash varies from 2.20 to 6.48 per cent, with an 

 average of 3.38 per cent. The lowest insoluble ash was obtained from 

 a cocoa treated with 2.5 per cent potassium carbonate and 10 per 

 cent water. The highest is that of a cocoa treated with 1.75 per cent 

 magnesium carbonate, which is insoluble in water. Aside from this 

 one, the maximum water-insoluble ash is 4.93 per cent. 



The alkalinity of water-soluble ash varies from 2.25 to 11.15 cc 

 of N/10 acid per gram. Naturally, those treated with ammonium 

 and magnesium carbonates show the lowest soluble alkalinity. Ex- 

 cluding these, the minimum is 4 cc. As might be expected, the maxi- 

 mum alkalinity is shown in those cocoas which have been treated 

 with sodium carbonate. The alkalinity of the insoluble ash shows 

 no points for discussion. The total alkalinity follows about the 

 same order as the soluble alkalinity and for approximately the same 

 reasons. 



The protein varies from 27.75 to 36.18 per cent. This latter figure 

 is attained on sample 26, which was treated with 4 per cent am- 

 monium carbonate. This would probably explain the high protein 

 percentage on the theory that a portion of the ammonia was not 

 volatilized. Excluding this sample, the maximum is 33.79 per cent. 



Table 4 shows the effect on the ash constants of the cocoa of the 

 addition of alkali, or the effects of various alkalies on the same blend. 

 Each pair was made from the same blend of beans and treated as 

 stated in Table 4. 



The results obtained are those which are normally expected on a 

 treated cocoa, namely, that the total water-soluble ash and the 

 alkalinity of the water-soluble ash are greatly increased, while the 

 water-insoluble ash is decreased. This proves to be the case in 

 every sample except sample 26, which shows practically no change 

 in the ash constants. It is quite evident that the treatment with 

 ammonia has made no change in the ash of the cocoa. Further 



