2 BULLETIN 666, U. S. DEPARTMENT OF AGRICULTURE. 



it to ferment and then stopping the fermentation by heat : or simply 

 adding water to the cocoa and heating the product to drive off this 

 added water, have all been employed. When alkalies are used, the 

 cocoa beans are partly roasted, then cracked, and the shells removed. 

 The nibs are then treated, either in the roaster or in a separate 

 kettle, with, the salt dissolved in water, or, in the case of magnesium 

 carbonate, made into an emulsion with water. The roast is then 

 finished, heat being applied gradually to drive off the water com- 

 pletely. Sometimes this procedure is varied by roasting the beans 

 entirely and cracking and winnowing them, adding the chemical 

 to the cracked nibs in a water-jacketed kettle and applying heat until 

 the product is dry. Although the first occasion for the employment 

 of any of these methods was undoubtedly an attempt to develop a 

 darker color in cocoas, it was later claimed that the alkali treatment 

 renders the cocoa more soluble in water. There is practically no 

 doubt that any of these treatments produces a cocoa which forms 

 a more perfect suspension hi the cup and which does not separate 

 as readily from the water solution as does an untreated cocoa. This 

 is largely due to the fact that a portion of the starch is gelatinized 

 by the action of the water in the treatment of the cocoa. It is also 

 claimed that the alkali disintegrates some of the cell walls, and in this 

 way enables the cocoa to stand up better in the cup. 



This investigation was undertaken primarily with a view to ascer- 

 taining whether the alkali treatment does hi fact render the cocoa 

 more soluble, and what changes take place in the cocoa as a result 

 of this treatment. Comparatively little work has been done on the 

 subject of alkali-treated or "Dutched' ; cocoas, and practically 

 none with the object of the investigation here reported. 



Farnsteiner * has made an exhaustive study of the effect on the 

 cocoa ash of the addition of varying amounts of potassium carbonate 

 and magnesium carbonate to cocoa. His work was carried out by 

 treating the raw cocoa bean, which had been shelled and ground as 

 finely as possible, with a definite amount of these reagents, ashing 

 the mixture, and determining the ash constants on this product. 

 In addition to this, he has worked out formulas for the determination 

 of the amount of alkali used. He also examined a large number of 

 cocoas in the same way, and endeavored from his researches to 

 determine the percentage and kind of alkali employed. Other refer- 

 ences to treated cocoas may be found in "Cocoa and Chocolate." 2 

 Practically the only subjects treated of in this book are the ash 

 constants and the method of manufacture of Dutch process cocoa. 



1 Z. Nahr. Oenussm. (1908) 16: 625-645. 



» Whymper. P. Blakiston's Son & Co. (1912): 103-110, 117, 231-238. 



