18 BULLETIN 666, U. S. DEPARTMENT OF AGRICULTURE. 
of alkali treatment as is the commonly accepted ratio of the water- 
soluble to water-insoluble ash on the cocoa; in fact, in samples such 
as 26, which was treated with ammonium carbonate, and 50, which 
was treated with magnesium carbonate, the ratio of the water- 
soluble ash to the water-insoluble ash of the water-soluble matter and 
also the alkalinity ratio show the presence of an added alkali where 
the ordinary ratio does not. 
ACTION OF VARIOUS ALKALIES ON COCOAS. 
In Tables 3, 6, and 9, the alkali-treated cocoas are listed according 
to the alkali used and according to the percentage of that alkali, 
making it possible to note the action of the various alkalies in their 
various proportions. Results are not exactly comparable, because 
the same blend of beans was not used in each case, but, as showing the 
general effect of each of these alkalies, and as comparing the effects of 
the various alkalies, they are of great value. Samples 34, 35,37, and 
38 are, as has been mentioned, imported cocoas, and the alkali used 
is unknown. Also, as has been pointed out, sample 2, which was 
supposed to have been made with 2.5 per cent potassium carbonate, 
contains less than this amount, judging from the analysis, probably 
as little as 1 per cent. The only points worthy of note in this con-. 
nection in Table 3 are: That the presence of sodium carbonate 
can be detected by the high alkalinities in practically every case; 
that ammonium carbonate exercises no effect on the normal cocoa 
ash, but that it is extremely difficult to remove all of the ammonia, 
as is shown by the fact that the indicated total protein on this sample 
is somewhat higher than normal; that the use of magnesium carbonate 
has the effect of increasing the water-insoluble ash and the alkalinity 
of the water-insoluble ash. In Table 6 the differences in the action 
of the alkali are more apparent. It will be noted that the treatment 
with 2.5 per cent potassium carbonate and 10 per cent of water gave — 
the maximum total soluble matter, the use of the water evidently 
increasing the solubility of the cocoa. A curious point in this con- 
nection is the fact that as the percentage of potassium carbonate 
used increases, the soluble matter decreases, with the exception of the 
one mentioned where water was also used. Therefore, if the intention 
is to obtain a more soluble cocoa, the use of 1 per cent potassium 
carbonate rather than larger amounts is to be recommended. This 
difference is still more apparent when the soluble matter minus the |; 
ash is calculated. Sodium carbonate has a much greater solvent 
action than has the potassium carbonate. Slightly less protein is |, 
dissolved by sodium carbonate than by potassium carbonate, but the 
total soluble matter, ash-free soluble matter, and ash- and protein-free 
soluble matter are greater in those cocoas which have been treated /}) 
with sodium carbonate. A mixture of the two, as shown in sample 
