COCOA BY-PRODUCTS 13 
engaged commercially in extracting cacao fat with a solvent, sent 
in samples. 
Certain analytical data, usually not considered by the fertilizer 
chemist, such as the contents of crude fiber and alkaloids, and the 
calorific value, have been determined on a number of the samples, 
because it was recognized that utilization of cocoa by-products as 
fertilizer material is in competition with their use as fuel, as feed 
stuff, and as a source of theobromine. 
The samples were ground to pass a 40-mesh sieve preparatory to 
the chemical examination. The methods of chemical analysis of 
the Association of Official Agricultural Chemists (3) were followed 
except as hereinafter noted. 
Total organic and ammoniacal nitrogen was determined by the 
Kjeldahl-Gunning-Arnold method in the nitrogen section of the 
Bureau of Chemistry. The determination of theobromine plus 
caffeine was made in the drug control and pharmacognosy labora- 
tories of the Bureau of Chemistry by a method worked out in those 
laboratories. Features of this method are: (1) Suspension of the 
cocoa material in weakly acidified water, and extraction of the 
alkaloids with chloroform in a special continuous-extraction appa- 
ratus; (2) purification of the extracted alkaloids by sublimation. 
It was found that higher yields of the alkaloids were obtained by 
this method than by the one-time official Dekker-Kunze method 
(J7, p. 25^-255) ; and the theobromine and caffeine recovered were 
of high purity. It was considered unnecessary to determine the 
bases separately, as the caffeine is known to amount to only about 
5 per cent of the total alkaloid content. 
The examinations relating to fuel value, involving determina- 
tions of moisture, volatile matter, ash, fixed carbon (by difference), 
sulphur, and calorific value, were made on other subsamples, in the 
coal analysis laboratory of the United States Bureau of Mines. 18 
In both the ether and petroleum-ether extractions it was found 
to be necessary to mix asbestos with the charges of material to 
render them more permeable to the solvent. The asbestos was first 
washed and ignited and its only effect on the determinations was 
to insure complete extraction. 
The method used in determining crude fiber differed from the 
one adopted as official for cacao products, in that purified and 
ignited asbestos was added to the charge to facilitate filtration; 
and linen, and asbestos felt in a Gooch crucible, were used as the 
filtering mediums after the acid and alkali digestions, respectively. 
Table 2 contains the analytical data obtained for the cocoa press 
cakes, including the sample of " whole bean " cake. 19 The results 
of the chemical investigation of the solubility and quality (or so- 
called availability) of the nitrogen of the press cakes, extracted 
cocoas, and cacao shells, are presented in Table 3 (p. 20). Table 4 
(p. 30) contains the results of the analyses of the solvent-extracted 
cocoas; and Table 5 (p. 37) shows the composition of the two 
samples of cacao shells. 
18 Those determinations woro made by IT. M. Cooper. The results: have been correlated 
With the moisture and ash data obtained on the original samples in this laboratory. 
19 Cheek determinations for a numl^er of the results were made by V. O. Lundstrom 
and R. M. Jones, assistant chemists, and W. O. Robinson, associate chemist, Bureau of 
Soils ; and M. R. Coe, assistant chemist, Bureau of Chemistry. 
